What Matched Betting Is, in One Paragraph
Matched betting is placing two bets that cancel each other out so the result no longer matters, and keeping the gap between the two prices as profit. You back a selection at a bookmaker at 2.00 and lay the same selection at an exchange at 2.02. If it wins, your back bet pays. If it loses, your lay bet wins. Either way you are up by the difference, and you have converted a free bet into cash with no exposure to the result.
That is the mechanism. Everything else in this guide is about the gap between what the arithmetic promises and what the bookmaker actually does about it, which is where the losses come from.
Two things to be clear about first. This is not betting: there is no view on the selection and no skill involved, it is a financial operation executed on a sportsbook and an exchange. And it is not long-term income, because the supply of welcome offers per human is finite. Our arbitrage guide covers the pure no-risk version of the same idea.
The Arithmetic: Back, Lay, Stake and Liability
You need four terms and they are easy to mix up.
A back bet is a normal bet. You stake a set amount at set odds, and win your stake times the decimal odds if the selection wins. Stake 100 at 2.00 and you win 100 profit, keeping your stake back.
A lay bet is a bet on the selection not happening. You are the bookmaker. You set a lay stake, the exchange sets the lay odds, and the exchange computes your liability: what you pay out if the selection wins. The relationship is the bit that catches everybody out. Liability equals the lay stake multiplied by (lay odds minus one).
Worked example, ignoring exchange commission. You back Team A at 2.00 for a stake of 100. You lay Team A at 2.02. For equal profit either way, the lay stake is 100 times 2.00 divided by 2.02 = 99.01, and the liability is 99.01 times 1.02 = 100.99.
If Team A wins: your back bet returns 200, profit of 100. Your lay bet loses its 100.99 liability. Net minus 0.99. If Team A loses: your back bet loses 100, your lay bet wins 99.01. Net minus 0.99.
So a perfectly matched pair at those prices loses 0.99. That loss is the qualifying bet, and it is the cost of claiming a bonus. The reason is the 1 per cent spread between the two odds. No arbitrage there, just a cost. The back/lay mechanics are set out in our lay betting guide.
Every Outcome of the Worked Example
The point of this table is that there is no third outcome. That is what makes the operation safe and what makes it boring.
| Outcome | Back bet at 2.00, stake 100 | Lay bet at 2.02, stake 99.01 | Net result |
|---|---|---|---|
| Team A wins | Wins 100 profit | Loses 100.99 liability | Minus 0.99 |
| Team A loses | Loses 100 stake | Wins 99.01 profit | Minus 0.99 |
| Draw in a 3-way market | Loses 100 stake | Wins 99.01 profit | Minus 0.99 |
| Market voided | Stake returned | Lay voided, no loss | Zero |
| With 2% commission on the lay win | Wins 100 | Wins 97.03 net of commission | Minus 2.97 |
| With 2% commission on the lay loss | Wins 100 | Loses 100.99, no commission | Minus 0.99 |
The Commission Detail That Costs People Money
Exchange commission is charged on net winnings, not turnover and not losses. On a matched pair it only bites on one of the two outcomes, which makes the result asymmetric and quietly erodes the profit.
Recompute at 2 per cent. When Team A wins, the back bet returns 200 and the lay costs 100.99, so minus 0.99. When Team A loses, the back costs 100 and the lay wins 99.01, from which the exchange takes 1.98, so minus 2.97. The worst case is now 2.97, not 0.99.
That matters for two decisions. The choice between a commission-charging exchange and a no-commission one can be worth more than the bonus on a small offer, so compare both. And if you lay at 2.02 and the selection wins, the liability is capped at what the exchange computed, so there is no commission on that side. The only real exposure in a matched pair is the qualifying loss, and commission is part of it. The broader mechanics are in our betting exchange guide.
From Qualifying Loss to Cash: A Real Example
Build a complete offer. A book offers a 50 free bet on a deposit, on any single at minimum odds of 1.50, with no wagering requirement on the free bet itself.
Stage one, the qualifying bet. Back at 1.80 for a stake of 50 to trigger the free bet. The lay stake is 50 times 1.80 divided by the lay odds. At lay odds of 1.84 that is 48.91, with a liability of 48.91 times 0.84 = 41.08. The qualifying loss is about 1.10 plus a little commission.
Stage two, the free bet. A free bet has no cash value of its own, only its winnings, so back at a short price where the winnings are likely to be lost and lay at a longer price. Back at 1.50 for a stake of 50, lay at 1.54. The lay stake for equal return is 48.70 and the liability is 26.30. If the selection wins you gain 25 on the back and lose 26.30, so minus 1.30. If it loses you lose nothing on the free bet and win 48.70 less 2 per cent commission, so plus 47.73. The worst case of the free bet is a loss of 1.30.
Total: about 1.10 paid on the qualifying bet, and 47.73 received on the free bet in the 71 per cent of cases it loses. Long-run expected value is about 33.5 against a real cash risk of about 2.40. That is roughly 33 units of profit on a 50 free bet, and it comes entirely from the price difference between two books, not from picking a winner.
The lower the back odds on the free bet, the better this gets, because a shorter price loses more often and the loss is capped at a small liability. That is the one optimisation that matters. Our free bets guide covers the same logic applied to bookmaker credits.
Seven Rules That Keep You Inside the Rules
None of this is legal advice and none of it is a suggestion. These are the boundaries, and crossing any one of them ends the activity.
- One account per book, ever. Unless the terms say otherwise in writing. Multi-accounting is the reason people lose balances they have already banked.
- Never deposit more than the offer requires. The free bet is 50 if you deposit 50. Depositing 500 to get 50 is a very expensive way to learn arithmetic.
- Withdraw early and often. A balance in a betting account is one the book controls. A balance in your bank account is money you have. Take it out before taking the next offer.
- Match the payment method. Deposit with the method you will withdraw to, in your own name. A matched-betting term of service is a small print document and it will list this. Our payment methods guide covers the routes.
- Bet at the market minimum odds, not the highest. You want the qualifying bet to settle fast, not to win.
- Do not assume a second promotion exists. Some books offer a second free bet to existing customers, usually at a worse price. Read the terms before assuming.
- Keep a ledger. One row per offer: book, offer, deposit, qualifying loss, free bet return, turnover completed, banked. Without it you will repeat offers and miscount your profit.
Where the Profit Comes From, and Why It Is Not Gambling
Worth being explicit, because "turning a free bet into 45 in cash" sounds like a trick and is not. It is a price discrepancy between two markets doing different jobs.
A bookmaker sets a back price at 2.00. An exchange, or another bookmaker, offers 2.02 on the opposite side. The difference compensates for two different businesses. The book is taking a position and managing a liability. The exchange is matching two traders and taking a 2 per cent cut. Each has a reason to shade the price in its own direction, and when they face each other the shades do not cancel, they leave a small cost for you.
That is a real arbitrage in the same way a market-making spread is a real cost, and the economics are set out in the arbitrage betting guide. The difference is that arbitrage survives indefinitely, because two prices genuinely differ on a genuine market, whereas matched-betting profit survives only until the bookmaker notices the pattern. That is also the answer to the question everyone eventually asks: books close matched-betting accounts not because the technique is cheating, but because a free bet is a marketing cost and a customer who converts it at 90 per cent efficiency is a customer that cost money. The book is not upset that you used the offer. It is upset that you used it perfectly.
Why the Supply of Offers Is the Real Limit
Forget the calculators. The binding constraint is how many welcome offers exist for you personally, and it is smaller than the marketing suggests.
There are roughly 200 to 300 betting sites offering something to a new customer, but the number offering a genuine matched-betting opportunity to someone in your situation is a fraction. Many offers are restricted to specific countries, many are net-wagering rather than free bets, many have a minimum deposit well above the free bet value, and a fair number are recycled under new branding.
Then the per-person filters bite. One account per person, enforced by name, address, date of birth, device fingerprint, bank account number, card number, IP address and payment method. A second account at the same book using the same bank account is trivially detected and is the same offence as multi-accounting regardless of intent. A second account at a different book using the same payment method is also linked in the background by data sharing, more often than people assume.
Finally the offers thin out through the year. The heavy spenders launch in January, February and September, and by November the market is scraped thin. So the honest expected figure: a complete beginner working steadily can realistically find 20 to 40 usable offers in a year and convert them into a few hundred currency units, while someone with real multi-book experience can do considerably better. Anyone telling you it pays a monthly salary is either sitting on a large account base or has not counted the time. Where to look first is in our bookmaker selection guide, though the only criterion that matters here is whether a new-customer offer exists.
The Risks Nobody Mentions
Four real risks sit underneath the arithmetic, and all four cost more than the qualifying loss.
Account restriction. The least bad outcome is a gub on your existing account: you keep the balance but cannot withdraw it. This happens when the book suspects multi-accounting rather than matched betting. There is no appeal, because appealing means admitting what you were doing.
Unmatched exposure. If you place the back bet and the exchange rejects the lay for any reason, a limit breach or a suspension, you are now holding a directional bet on a football match. This is the risk that turns a small controlled activity into a large loss. Lay first where the liability is the bigger number, or check the exchange limit accommodates your liability before you back.
Voided or dead-heat markets. Reduced stake, abandoned games and dead heats settle in ways that are not what you assumed. Lay bets on Asian handicap markets in particular void rather than refund, so if the underlying back bet pushes you can be left naked. Stick to simple two-way markets with clean settlement rules.
Turning it into a habit. This is the one I am most serious about. Matched betting is designed to be a one-off. Doing it monthly turns a bounded activity with a known endpoint into an open-ended one, and it encourages spending time on promotions instead of the thing you are reading guides to improve. Pick a weekend, work the offers, bank the money, and stop. Odds boosts are the specific trap, because an enhanced-odds offer is usually a matched-betting offer with a better price attached, and the better price does nothing unless you lay the other side. That is in our odds boosts guide.
A Sensible Plan for Your First Weekend
Do it once, properly, then decide whether to continue.
Saturday morning. Open a betting exchange and complete verification properly, because an unverified account gets limits cut. Have a calculator or spreadsheet with three functions: find lay stake, find lay odds from a back price, and find lay odds from a target liability. The first and last are not the same function and mixing them up is the classic error.
Saturday afternoon. Pick six books from your own list, based on offers that exist rather than reputation. Check the qualifying terms for each: minimum odds, minimum deposit, market restrictions, whether the free bet is stake-not-returned, and whether turnover applies.
Saturday evening. Work them in order. For each: deposit the minimum, place the qualifying bet at the minimum permitted odds on a market that settles within a few hours, match it, wait for settlement, then take the free bet at a short back price with a longer lay price. Bank each profit as you go rather than leaving it in the account.
Sunday. Do nothing. If six offers produced 250 in profit for about four hours of work, you now have a real number for what your time is worth, and you can decide about the next twenty offers with actual information rather than a forum post. Everything here assumes you are doing this yourself, with your own money, on your own accounts. Running it for other people, for a fee or a cut, is unlicensed gambling. Our responsible betting tools guide covers the account limits worth setting, because a promotional offer is still a bet.
