Arbitrage Betting

Arbitrage betting, often shortened to “arbing,” is a technique where a bettor places wagers on every possible outcome of an event across different bookmakers, using odds that are mismatched enough to lock in a profit no matter which outcome actually happens. It isn’t a prediction strategy at all — you don’t need to know which team will win, because you’re betting on all of them at once. The profit comes purely from inefficiency in the market, not from skill in reading form or statistics.

The mechanism relies on the fact that different bookmakers price the same event slightly differently, and occasionally those differences overshoot the built-in margin (the “overround”) that normally guarantees the bookmaker a profit whatever happens. When the combined implied probability of all outcomes, taken from the best available price at each bookmaker, drops below 100%, an arbitrage exists. Stake proportionally across those prices and you come out ahead regardless of the result.

This only works because bookmakers set odds independently and sometimes disagree sharply — one might be slow to move a price after team news breaks, another might be trying to attract action on an unpopular side, or a smaller regional bookmaker simply prices a market wrong. These gaps are usually small and short-lived, and finding them by hand across dozens of bookmakers is impractical, which is why arbers tend to use odds-comparison software that scans for them continuously.

It’s worth being clear that arbitrage betting is mathematically riskless but not practically riskless. Bookmakers actively limit or close accounts that show consistent arbing patterns, prices can shift or get suspended between placing your first and second leg, and stake limits on the “soft” side of the arb can cap how much profit you can actually extract.

Example

Take a Bundesliga match, Bayer Leverkusen against Werder Bremen, with a straightforward two-way market once you strip out the draw (say a “to qualify” tie in a cup replay, so only Leverkusen or Bremen can advance). Bookmaker A prices Leverkusen at 2.05 to advance. Bookmaker B, slower to react to Bremen’s makeshift back line, prices Bremen at 2.15.

Check the implied probabilities: 1 ÷ 2.05 = 48.78%, and 1 ÷ 2.15 = 46.51%. Add them together and you get 95.29% — comfortably under 100%, which confirms an arbitrage exists worth roughly 4.7%.

Now work out the stakes. Say you want to commit a total of €500. Stake on Leverkusen = (€500 × 48.78%) ÷ 95.29% = €255.90. Stake on Bremen = (€500 × 46.51%) ÷ 95.29% = €244.10.

Check both outcomes:

• Leverkusen advances: €255.90 × 2.05 = €524.60 returned.
• Bremen advances: €244.10 × 2.15 = €524.82 returned.

Both returns land close to €524.70 (the small 22-cent gap is just rounding), against a €500 outlay — a guaranteed profit of about €24.70, or 4.9% on turnover, whichever team goes through. That’s the whole trick: size each stake so every branch pays out the same amount.

Key Points

• The overround has to invert: an arbitrage only exists when the sum of implied probabilities (1 ÷ decimal odds, added across every outcome from the best-priced bookmaker) falls below 100%. Above that, you’re just backing multiple outcomes at a guaranteed loss, which is how most people wrongly think they’ve found one.
• Stake proportionally, not equally: splitting €500 evenly between Leverkusen and Bremen in the example above would leave one outcome paying more than the other. Weight each stake by its own implied probability divided by the total, as shown, so every result returns the same amount.
• Margins are thin, so errors are expensive: a 2-3% edge disappears fast if you fat-finger a stake, misread a market (e.g. one site includes overtime, the other doesn’t), or a price moves before your second bet lands. Double-check market rules match exactly across both bookmakers before staking.
• Accounts get restricted: bookmakers monitor for arbing signatures — mismatched stake sizes, betting only on the value side of two-way markets, no losing bets ever. Persistent arbers typically see stake limits cut or accounts closed, which caps this as a long-term strategy at any single bookmaker.
• Speed and liquidity matter more than the maths: the arithmetic is trivial; the hard part is placing both legs before either price moves or the bookmaker suspends the market, and finding enough stake room on the smaller book to match what the bigger one allows.
• It’s a market-inefficiency play, not a form-study play: unlike value betting, you gain nothing from expertise in the sport itself. The edge lives entirely in price discrepancy, so time spent is better spent scanning odds feeds than watching match footage.