Why Betting Systems Fail: The Truth About Martingale and Other Staking Plans

Every weekend, somebody somewhere is sure they've cracked it. There's a forum post with a screenshot of a rising balance, a PDF promising a "secret staking plan", a mate who swears that doubling up after every loss has never let him down. The appeal is obvious: who wouldn't want a formula that turns a profit on demand?

Here's the awkward bit. No betting system has ever beaten the house edge. Not the Martingale, not the Fibonacci, not the Labouchere or the D'Alembert. They've been around for centuries, and casinos and bookmakers are still very much in business, because changing the size of your stake can't change what each bet is worth.

So why do these systems feel so convincing? We'll look at how they work, where they break, the psychology that keeps people believing, and what a smarter approach looks like.

Before you can see why systems fail, you need to know what you're up against. Every price a bookmaker or casino offers has a little built in for the house.

Take a coin toss. Fair odds would be 2.00 on heads and 2.00 on tails. A bookmaker might offer 1.91 on each side instead. Work out the implied probability of each (1 divided by 1.91, or 52.36%) and the two add up to 104.71%. That extra 4.71% is the overround, sometimes called the vig, and it's the bookmaker's margin on every bet you place. It varies from firm to firm and market to market: sharp operators like Pinnacle and the betting exchanges price tighter than most high-street shops, so the same event can carry a noticeably different margin depending on where you bet.

Casino games are even more blunt. European roulette has a house edge of 2.7% (a return to player of 97.3%), while American roulette, with its extra double zero, carries 5.26%. That edge applies to every spin, whatever happened on the last one. It's also why multi-bets sting so much: every extra leg stacks another margin on top.

The Martingale is the most famous staking system going, and probably the one that has emptied the most wallets. Its appeal is its simplicity. After every loss you double your stake; after a win you go back to your starting bet. The theory says that when you finally win, you recover everything you've lost and pocket a profit equal to your first stake.

Run it with a £1 starting bet and it looks lovely: lose £1, lose £2, lose £4, then win with £8. You've staked £15 in total, got £16 back, and you're £1 up. Foolproof, isn't it? Eventually you must win, and when you do you're ahead by one unit. So what could possibly go wrong?

Three things bring it down, every single time. The first is that the stakes explode. Each loss doubles the bet, so the money you have to risk races far ahead of the £1 you're trying to win:

Ten losses in a row means a £1,024 bet to win back a £1 profit. At that point the risk-to-reward ratio has stopped being a strategy and become a dare.

Your bankroll runs out

Start with £1,000 and a £1 base bet and you can absorb nine losses in a row (£511 gone, £489 left). The tenth would need a £512 stake you can no longer cover. In European roulette, nine straight losses on an even-money bet turn up roughly 0.25% of the time (about 1 in 400 starting points), and ten about 0.13% (roughly 1 in 780). Those sound like long shots, but over thousands of bets they stop being rare.

Limits cap the climb

Casinos and bookmakers set maximum stakes. Once the bet your system demands is above the limit, you can't follow the progression any further, and the losses you've already taken stay exactly where they are.

More often than most people think. Take European roulette and an even-money bet, where each spin has a 19/37 chance of losing. The chance of at least one run of ten straight losses climbs steadily with the number of spins (this is straight probability maths, assuming independent spins). Play long enough and the unlikely becomes routine:

The Martingale has plenty of cousins. They dress the same idea up differently, and they all end up in the same place.

Fibonacci

This one uses the famous sequence (1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89...) to set your stake after each loss, and after a win you step back two places. Because it climbs more slowly than doubling, it feels safer. But recovery is slower too, and stakes still balloon in a long bad run. Meanwhile the expected value doesn't budge: the house edge applies to every bet in the sequence.

Labouchere

Also known as the cancellation system. You write down a line of numbers such as 1-2-3-4-5-6 and bet the sum of the first and last. Win, and you cross both out; lose, and you add the amount you staked to the end of the line. Cross out every number and you've banked a profit equal to the original line's total, which is £21 on £1 units. It feels wonderfully sophisticated, but every loss lengthens the line and raises the next stake, so a bad run compounds just as it does with the Martingale. With a negative expected value on each bet, the line closes less often than the rules suggest, and a long losing run can leave you staking far more than you ever planned.

D'Alembert

The "gentle" option: add one unit after a loss, take one off after a win. If you win and lose equally often, you come out ahead. The catch is that you don't win equally often. On a European roulette even-money bet you win only 48.65% of the time, so stakes creep up and clawing back from a bad patch is painfully slow. It's the tamest of the lot, but it still loses over the long run.

Here's the principle that sinks every system: no staking pattern can change the expected value of a bet.

Take European roulette. Every £1 on an even-money bet is worth, on average, a loss of about 2.7p: (18/37 × £1) + (19/37 × −£1) = −£0.027. Stake £2 and the expected loss doubles to 5.4p; £4 costs 10.8p and £8 costs 21.6p. Spread £1, £2, £4 and £8 across four spins and your expected loss is about 40.5p, exactly what you'd lose on average by putting £15 on a single spin. Split your money up or bet it in one go, the maths can't tell the difference.

Plenty of systems tell you to wait for a streak before you start, on the logic that after several losses a win must be due. That's the gambler's fallacy, and it's one of the costliest thinking errors in betting.

Every spin, every coin flip and every match is independent. Throw six heads in a row from fresh flips and you've pulled off a 1-in-64 shot. But the chance of heads on the sixth flip, after five heads already, is still exactly 1 in 2. The roulette wheel has no memory, and the coin doesn't know what it landed on last time.

The most famous example comes from the Monte Carlo casino in August 1913, when black reportedly came up 26 times in a row. Gamblers piled onto red, convinced it was overdue, and many lost heavily. On every one of those spins the chance of red was still 18/37.

Variance is what keeps these systems alive. A Martingale player is usually ahead after a short session, because most sequences end with a small win, and that breeds dangerous confidence.

Give it time, though, and the law of large numbers goes to work. The longer you play, the closer your results drift towards the mathematical expectation. You're swapping lots of tiny wins for the odd enormous loss, and sooner or later that big one lands and wipes out weeks of patient profit. It can happen in minutes.

If betting systems are doomed, why do so many people swear by them? A handful of very human quirks do most of the work.

Selective memory

We replay the winning sessions and shrug off the losses. Picture ten Martingale sessions: eight tidy little wins and two disasters. Most people remember the eight and blame the two on bad luck or a slip, never on the system itself.

Chasing losses

After a bad run the urge to win it back is overpowering, and doubling up feels rational when you're already deep in. That's the sunk cost fallacy at work, and it leads to ever-bigger bets and eventual ruin, which is exactly what these systems encourage.

Seeing patterns that aren't there

Humans evolved to spot patterns because it kept us alive. With random events it misfires: we see hot streaks and corrections that aren't real, and then chase them.

The small-win illusion

Because most sessions finish a little ahead, it's easy to miss that you're trading lots of small wins for the occasional huge loss. The system works, right up until it doesn't.

Not all betting maths is hopeless. The Kelly criterion is a genuinely useful tool, but only in one situation: when you already have an edge.

The formula is f = (bp − q) / b, where f is the fraction of your bankroll to stake, b is the decimal odds minus 1, p is your probability of winning and q is your probability of losing (1 − p). Say you have a 60% chance on an even-money bet (decimal odds of 2.00). Then f = (1 × 0.60 − 0.40) / 1 = 0.20, so Kelly says to stake 20% of your bankroll.

Here's the catch. Kelly tells you how much to bet once you have an edge; it can't create one. Feed it a game with negative expected value and the answer is to bet nothing at all.

Even with a genuine edge, Kelly is harder to follow than it looks. In a well-known experiment by Haghani and Dewey (analysed at gwern.net), 61 quantitatively trained players were given $25 and 30 minutes to bet on a coin that landed heads 60% of the time, with winnings capped at $250. It was about as good a bet as you could ask for. Yet the average player finished with just $91, and roughly three in ten went bust.

The lesson is that having an edge is only half the battle; staking it sensibly is the other half. Full Kelly is a rollercoaster, which is why many serious bettors prefer a fraction of it, or simply compare it with flat staking to find a level they can stick to.

If systems are worthless and the house always has the edge, is profitable betting impossible? Not quite, but it takes a completely different mindset.

Value betting means backing a selection when the bookmaker's odds imply a lower chance than the one you believe the outcome really has. That gap is what creates positive expected value. Say a bookmaker offers 2.00 (an implied 50%) on something you rate at 60%. Your expected return on a £1 stake is (0.60 × £1) + (0.40 × −£1) = +£0.20. That's a +EV bet, and it's the only kind worth making. For a football-specific walkthrough, try our guide to finding value bets in football.

Knowing who you're betting with helps you spot value. Sharp bookmakers, with Pinnacle the usual example, court professional bettors, run tight margins, move their odds fast and are the best guide to a fair price. Soft bookmakers such as Bet365 and Ladbrokes cater for casual customers, carry wider margins in many markets and can be slower to adjust in some of them, which leaves more pricing gaps. That makes value easier to find at the soft end, though "easier" is relative.

Why value betting is hard

Be realistic about what it takes. You need to estimate probabilities better than the bookmakers, whose whole business depends on getting them right. You need specialist knowledge of particular leagues or markets, and a willingness to look where bookmakers have less information and update more slowly. And you need to record every bet to prove your edge is real, with the discipline to bet only when the value is there.

If you'd rather build a proper approach than chase a miracle system, start here.

Learn probability and expected value

Every bet has an expected value. If you can't estimate it, you're guessing. Get comfortable converting odds into implied probabilities and comparing them with your own numbers.

Specialise

Pick leagues, teams or markets where you can genuinely know more than the average bettor. Plenty of professionals focus on corners where bookmakers have less information, such as lower-division football or women's sport.

Track everything

Log every bet: the odds, your estimated probability, the result and your profit or loss. After a few hundred bets you'll have real evidence of whether you have an edge or have simply been riding variance.

Stake with fractional Kelly

If you do have a genuine edge, a half or a quarter of the Kelly stake trims the swings while still growing your bankroll. Our bankroll management guide covers the practical side.

Pass when there's no value

The hardest skill in betting isn't placing the bet, it's walking away. If you can't see clear value, don't bet.

Know what you're paying for

Betting for fun is perfectly fine, but treat it as entertainment. Set firm loss limits and never stake more than you can afford to lose.

Spot any of these and you could save yourself a lot of money:

• "Guaranteed" profits. Nothing in betting is guaranteed, and a promise of certainty is a mathematical impossibility.
• Doubling up after losses. Every Martingale variant ends in ruin eventually.
• Claims to "beat the house edge" with staking. No staking pattern can touch the edge.
• Pages that push a system and a sign-up link together. Ask who gets paid when you join.
• Systems that ignore limits and bankrolls. That is exactly where they fail.
• Testimonials with no long-term verified record. A good fortnight proves nothing.

Betting systems have lasted for centuries because they scratch a psychological itch, not a mathematical one. They give you a feeling of control, serve up plenty of little wins to keep the faith alive, and put off the reckoning with clever staking. But maths isn't fooled.

The house edge sits inside every bet you place, and no progression, sequence or cancellation method can move it. The Martingale, Fibonacci, Labouchere and D'Alembert are all equally worthless in the long run. Our view? Skip the system, learn to spot value, and use staking to protect your bankroll rather than to try to outsmart the bookies.

Value betting is hard, slow and comes with no guarantees. But at least it stands on solid mathematical ground, which is more than any doubling plan can say.

• Value Betting Football: How to Find +EV Bets - Learn the only mathematically sound approach to profitable betting
• Why Most Parlays Lose: The Math Behind Parlay Betting - See how the house edge compounds in multi-bets
• Flat Betting vs Percentage Betting - Compare two popular staking methods
• Bankroll Management in Betting: The Complete Guide - Protect your betting funds with proper staking

18+. Please gamble responsibly. For free, confidential support visit BeGambleAware.org.