Arbitrage Betting Explained: How Sure Bets Work

One bookmaker has Player X at 2.10 (11/10) for a tennis match. Another has Player Y at 2.05 (21/20). These are illustrative prices, but the idea is real: split £100 between them the right way and you collect about £103.73 whoever wins. No form study, no hunch about the second set. That's an arbitrage bet, or "arb", and it's the betting version of buying cheap in one shop and selling dear in the next.

Arbers call them sure bets. We'd call them conditional bets. The return is only locked in if both bets are accepted at the prices you saw and settled the same way, and most of the craft is about protecting that "if".

We'll start with the one-line test that spots an arb and the formula that sets your stakes. Then come worked examples for two-way, three-way and exchange arbs, the main types of arbitrage betting and an honest look at what goes wrong. After that we cover where arbing stands with UK law and tax, and finish with a pre-bet checklist.

Every bookmaker prices its markets with its own model and builds in its own margin. Add up the implied probabilities on one firm's market and you'll land somewhere over 100%, say 105%. That extra slice is the bookie's edge (our guide to bookmaker margin and fair odds shows how to measure it), and on a single firm's prices it makes an arb impossible.

Things change when you take the best price for each outcome across the whole market. Different firms read the same event differently, and their margins sit in different places. Every so often the top prices, each from a different firm, add up to less than 100%. That's your arb.

What opens the gap

• News breaks. OddsMonkey points to team or player news, and to bookies that are simply slow to react when a market moves. One firm adjusts while another is still catching up.
• Honest disagreement. Two pricing teams can study the same match and land on different numbers.
• Competitive pressure. A firm chasing business can push a headline price above the rest of the market.
• Mistakes. People and software both misprice things. Errors throw up the juiciest-looking arbs, and the riskiest.

Spotting a gap is only half the job. The maths is what tells a real arb apart from a price that just looks generous.

Arb maths runs on decimal odds, so convert any fractions first: divide the top number by the bottom and add one. 11/10 becomes 2.10, and 5/2 becomes 3.50. Then run the test.

Arb % = (1 / odds A) + (1 / odds B) [+ (1 / odds C) ...]

Under 100%   -> arb: the same return whatever wins
100% or more -> no arb

If the prices pass, each outcome takes a share of your total stake in proportion to its implied probability. That share is what evens out the return.

Stake on outcome = Total stake x (1 / outcome odds) / Arb %
Return (any result) = Total stake / Arb %

Worked example: a two-way tennis arb

These are illustrative odds, not live prices.

Player X at Bookmaker A: 2.10 (11/10) -> 1 / 2.10 = 47.62%
Player Y at Bookmaker B: 2.05 (21/20) -> 1 / 2.05 = 48.78%
Arb % = 96.40%

Total stake: £100
Stake on X = 100 x 47.62 / 96.40 = £49.40
Stake on Y = 100 x 48.78 / 96.40 = £50.60

X wins: 49.40 x 2.10 = £103.74
Y wins: 50.60 x 2.05 = £103.73
Profit: about £3.73 (3.7%)

The shorter price, 2.05, gets the bigger stake because it needs more money to reach the same return. Now look at the size of the prize. Two near-even prices produced a profit under 4%. That's what a healthy arb looks like.

If you've read our guide to dutching and how to dutch a bet, this will feel familiar. A bookie-v-bookie arb is really a dutch across two firms where the best prices happen to add up to less than 100%.

You don't need two prices near evens. Here's a lopsided two-way market, again with illustrative odds.

Favourite at 1.20 (1/5): 1 / 1.20 = 83.33%
Outsider at 8.00 (7/1):  1 / 8.00 = 12.50%
Arb % = 95.83%

£100 total: £86.96 on the favourite, £13.04 on the outsider
Return either way: about £104.35 (profit about £4.35)

Most of the money sits on the favourite, with a small stake covering the upset. You get the same return either way, and the profit is still in low single figures.

Adding the draw

Football's match result market has three outcomes, so a football arb needs three legs at three bookmakers. The odds below are illustrative too.

Home win at Bookmaker A: 2.80 (9/5)  -> 35.71%
Draw at Bookmaker B:     3.60 (13/5) -> 27.78%
Away win at Bookmaker C: 3.20 (11/5) -> 31.25%
Arb % = 94.74%

£100 total: £37.70 / £29.32 / £32.98
Return whatever happens: about £105.55 (profit £5.55)

On paper the football arb looks better. In practice you now need three firms to accept three bets at the prices you saw, which triples the execution risk. OddsMonkey flags football arbs as riskier because of the chance of a draw, which is what forces that third leg.

Our view: learn on two-way markets first. Tennis and other sports without a draw keep the moving parts to a minimum.

This is the arb most UK bettors meet first. Outplayed calls it "the most popular form of arbitrage betting in the UK", and Livescore calls it the most common type. You back a selection with a bookmaker and lay the same selection on an exchange. Ignore commission and you have an arb whenever the back odds beat the lay odds.

Lay stake = Back stake x Back odds / (Lay odds - commission rate)
Lay liability = Lay stake x (Lay odds - 1)

Say you back £100 at 2.30 (13/10) and lay at 2.10 (11/10). Again, the odds are illustrative.

No commission:
Lay stake = 100 x 2.30 / 2.10 = £109.52
Liability = 109.52 x 1.10 = £120.48

Selection wins:  £130.00 back profit - £120.48 liability = £9.52
Selection loses: £109.52 lay win - £100.00 back stake = £9.52

With 2% commission on the lay win:
Lay stake = 100 x 2.30 / (2.10 - 0.02) = £110.58
Profit either way: about £8.37

Commission is part of the price

Betfair Exchange only charges commission on your net winnings on a market, and nothing on a net loss. Your rate depends on your rewards package, the Market Base Rate and any discount you have. Betfair's own worked example uses 2% for its Basic package, so £100 of net profit costs £2. The gap between the back and lay prices has to cover that before you've got an arb.

Then there's the cash. Laying £109.52 at 2.10 needs £120.48 of liability sitting in the exchange account, on top of the £100 at the bookmaker. Arbing ties up far more money than the profit figure suggests.

Arbs come in a handful of shapes, and each one has its own catch.

Sharbing (shop plus arbing) works because high-street bookmakers update their prices more slowly than online books and run their own in-shop promotions, as OddsMonkey notes. A price on the counter can sit above the exchange lay for a while.

We'd give exchange-v-exchange arbing a miss, and Outplayed strongly advises against it too. Bookie v exchange is the natural place to start. Only one bookmaker has to take your bet, and the exchange shows you exactly what the lay will cost.

An arb is only locked in once every leg is accepted, settled under the same rules and paid.

The maths is the easy bit. Here's what catches arbers out.

The execution gap

OddsMonkey puts it well: the hard part "isn't usually finding an opportunity – it's executing both bets before the odds change. Even a few seconds can remove the profit margin." Get your first leg on, find the second price has dropped, and you're left with a thinner arb, a losing one or a straight bet. Our habit is to place the leg most likely to move or be cut first.

Palpable errors and voids

Nearly every bookmaker reserves the right to void bets struck at an obvious pricing error, known as a palpable error, or to resettle them at the corrected price. Run that through our tennis example. If the £50.60 leg at 2.05 is voided, you're holding £49.40 on Player X at 2.10. A win pays £54.34 profit and a loss costs you £49.40. Your "sure bet" is now a straight bet.

That's why size is a warning sign. Betsmart doesn't mince words: "The biggest arbs on any scanner are often errors." It puts a normal arb at 1% to 3% and suggests checking anything above about 5% against a third bookmaker. bettingexpert says a return above 3% "will be a rare event". So when a two-way tennis arb shows 12%, treat it as a probable error, not a windfall.

If you think a void was unfair, start with the bookmaker's complaints process. If that doesn't settle it, you can take it to IBAS, the Independent Betting Adjudication Service.

Different settlement rules

Your two legs can play by different rules. Betsmart's tennis example is the classic: bookmakers settle retirements differently, so one leg can be voided while the other loses. Before you bet, read each leg's rules on retirements, extra time, dead heats and void conditions, plus Rule 4 if it's racing.

Yes, arbing is legal in the UK. The law regulates operators, not how customers pick their bets. Betsmart draws a useful line between what's legal and what's allowed. Your account is a contract under the bookmaker's terms, and those terms let firms cut stakes, exclude you from promotions or close accounts at their discretion.

Betsmart again: "A book does not have to prove you are arbing to limit you. Patterns that look professional are enough." That side of arbing deserves a guide of its own, and we've written two: how bookmakers detect arbitrage betting and do bookmakers limit winning players.

Tax

The Gambling Commission is clear that gambling winnings are tax-free for customers in the UK, whatever the amount. The tax falls on operators instead, through their Gross Gambling Yield. Our guide to tax on betting winnings in the UK covers where professionals can cross the line.

Promotions

Using arbs to take welcome offers and free bets is a separate matter. Promotions carry their own terms, so read them before you pair a bonus with an arb.

Two terms crop up in every arb. Backing is betting that something will happen, the normal bet you strike with a bookmaker. Laying, on an exchange, is betting that it won't. You act as the bookmaker, and your liability is the lay stake multiplied by the lay odds minus one.

A normal market adds up to more than 100%. An arb only exists when the best prices from different firms add up to less. Run these checks before any money goes down.

Once you've rounded your stakes to the penny, recheck the return on each leg. On an arb worth a few per cent, a rounding slip on the wrong leg can eat most of the profit.

• Respect the small numbers. A healthy arb returns a few per cent. Arbing is slow, steady work, and the biggest numbers on the screen are the ones most likely to vanish.
• Start two-way. Tennis and other no-draw sports give you two legs at two firms, and far fewer ways for things to go wrong.
• Bookie v exchange is the best classroom. Only one bookmaker has to accept the bet, and the lay maths teaches you how commission and liability really work.
• Price commission in from the start. Work out the lay stake with your commission rate built in, not bolted on afterwards.
• Fund both sides first. An arb you can't complete because the exchange balance is short is just a bet.
• Read the rules on every leg. Retirements, extra time and dead heats are where a neat pair of bets comes apart.
• Doubt any arb that looks too good. Above about 5%, check a third bookmaker before you go near it.

Set a bankroll you're happy to have tied up across several accounts, and keep your stakes inside it. However tidy the maths, an arb can still lose money.

• Dutching explained: calculator, strategy and how to dutch bet
• How bookmakers detect arbitrage betting: inside the high-tech arms race
• Do bookmakers limit winning players? Account restrictions explained
• Bookmaker margin and fair odds: how to remove the vig
• Lay betting strategy: how to trade on Betfair like a professional

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