The principle
A surebet (or arbitrage, "arb") exists when the best odds for each outcome, taken at different bookmakers, have implied probabilities that add up to less than 100%. In that case, by splitting the stakes correctly, you win whatever the outcome.
Example on a tennis match: player A is priced at 2.10 at one bookmaker, player B at 2.05 at another. Sum of the inverses: 1/2.10 + 1/2.05 = 0.476 + 0.488 = 0.964. That is 3.6% short of 100%: it is your guaranteed profit, 3.7% of the total stake (1/0.964 − 1).
Splitting the stakes
For an identical profit whatever the outcome, each stake is proportional to the inverse of its odds:
Stake on outcome i = Total × (1/cᵢ) / Σ(1/cⱼ)
On €100: €49.40 on A at 2.10 and €50.60 on B at 2.05. If A wins, you collect 49.4 × 2.10 = €103.70; if B wins, 50.6 × 2.05 = €103.70. The arbitrage calculator does the split for two or three outcomes and shows the profit.
Why they exist
Bookmakers don't align instantly. When news moves the market — an injury, a wave of money on the exchange — each bookmaker reacts at its own pace. For a few minutes, one is still quoting the old price while the other already has the new one: that is where most surebets are born. So they are rare (a few percent of markets), brief and often narrow (1 to 3%).
The real risks
A surebet is only sure on paper. Four things go wrong in practice:
- The odds move between your two bets. You place the first stake, the second price has dropped: you are now holding a single bet, not an arbitrage. Rule: place the first stake at the bookmaker that is slowest to correct.
- A bet is voided. Rules differ between bookmakers (postponed match, retirement in tennis, overtime in basketball): one leg is refunded, the other stays open. Check that both bookmakers apply the same rules on that market.
- Palpable error. Obviously wrong odds can be voided after the fact by the bookmaker. Surebets that are "too good" (10% and more) are often of this kind.
- Account limiting. This is the structural limit: mainstream bookmakers spot arbers and cut their maximum stakes to a few euros. A "limited" account is no longer any use for arbitrage, nor for value bets. The bookmakers that don't limit (Pinnacle, the exchanges) rarely offer odds above the market.
Where arbitrage fits
Arbitrage is an excellent school: it forces you to compare odds, understand the margin and act fast. But as a source of income, it hits a ceiling quickly: the profits are small, the accounts get closed, and the time spent is considerable.
Value betting is the logical next step. A surebet is in fact the meeting of two bets, at least one of which is a value bet; instead of covering both sides for a guaranteed 2%, you only play the side that has value — with variance, but with a far higher expectation, and without doubling your stakes. Proba Bet shows both: the risk-free combinations in their own tab, and the value-bet selection, graded A, B or C by reliability.
Key takeaways
A surebet is a gap between bookmakers, not a gift: it is rare, short-lived, and the accounts that abuse it get limited. Use it to learn to read the market, not to make a living from it. The real income is the fair odds, and the bet on the side where they are beaten.