Arbitrage and Sure Bets: How They Work and How They End
The arithmetic behind a sure bet, how the stake is split so both outcomes pay the same, what the return really is on turnover, everything that eats it — moving prices, stake limits, restricted accounts, voided legs and differing rules — and the hedging technique that outlives arbitrage.
A sure bet exists when two operators disagree enough about the same match that you can back every outcome and still be left with money. The arithmetic is real, it takes a minute to verify, and the return on a good one is under 4% of what you stake. Everything difficult happens after that calculation: the second price moves, the stake is capped, the account is restricted, a leg is voided. This page does the sums, then the subtraction — part of the wider map of betting strategies.
In short.
- An arb exists when the implied percentages of every outcome add to less than 100% across two operators.
- Split the stake in proportion to those percentages and both results pay the same amount.
- A 3.7% arb on ₦500 pays ₦19. The return is a percentage of turnover, and turnover is limited by capital and by how fast bets settle.
- Price moves, stake caps, restrictions, voided legs and mismatched rulebooks subtract from that ₦19; one void costs about fourteen good arbs.
- The same arithmetic used to close a position you already hold is more useful to most punters than the arbitrage itself.
When two prices leave a gap
Convert each price into the percentage it implies by dividing 1 by it, then add the outcomes. Within one operator that sum is always above 100%, and the excess is the mark-up described in the guide to margin. Across two operators it can occasionally fall below.
Operator A prices Over 2.5 goals at 2.10; operator B prices Under 2.5 at 2.05. That is 1 ÷ 2.10 = 47.62% and 1 ÷ 2.05 = 48.78%, adding to 96.40%. The missing 3.60% is the gap. Your return on the money you put through is 1 ÷ 0.9640 − 1 = 3.74%, and it arrives whichever way the match goes.
The two selections must cover every outcome exactly once. Over and Under on the same line qualify; a home win at one operator and a draw at another does not, because the away win is left uncovered.
Splitting ₦500 so both outcomes pay the same
Example. You have ₦500 to put through. Each leg takes the share of the total that its own percentage represents: the Over leg gets 500 × 0.4762 ÷ 0.9640 = ₦247, the Under leg 500 × 0.4878 ÷ 0.9640 = ₦253. If the match goes over, 247 × 2.10 = ₦518.70 comes back. If it goes under, 253 × 2.05 = ₦518.65. Profit is ₦18.70 either way, on ₦500 staked across two apps.
Two details survive contact with a real slip. Rounding both stakes to ₦250 changes the outcome to ₦525 one way and ₦512.50 the other — still profitable, no longer level. And ₦18.70 is the entire reward for funding two accounts, checking two rulebooks and waiting for settlement.
The return is a percentage of turnover
Arbitrage has no win rate; it has a throughput. What you earn is the margin multiplied by the money you push through, and that money is locked until both legs settle.
| Turnover on one arb | At 3.7% | At 2.0% | At 1.0% |
|---|---|---|---|
| ₦500 | ₦19 | ₦10 | ₦5 |
| ₦2,000 | ₦74 | ₦40 | ₦20 |
| ₦10,000 | ₦370 | ₦200 | ₦100 |
| ₦40,000 | ₦1,480 | ₦800 | ₦400 |
Read the first row against the last. The method is indifferent to your opinion about football and entirely dependent on capital, which is the opposite of every other approach in this section.
Everything that eats the margin
| What happens | What it does | Cost on the ₦500 example |
|---|---|---|
| The second price moves before you place it | Under drifts from 2.05 to 1.92 while you switch apps: the sum becomes 99.70% | Profit falls from ₦18.70 to ₦1.49 |
| It moves further | At 1.85 the sum is 101.67% and the arb is gone | Complete it for a certain −₦8, or hold ₦247 unhedged |
| The stake is capped | The mispriced side accepts ₦2,000, the other ₦50,000 | The whole arb caps at about ₦4,050 of turnover |
| The account is restricted | Maximum stake cut to a fraction, usually on the account you need most | Capacity falls without the margin changing |
| One leg is voided | Abandonment or a rule that returns the stake on one side only | The other leg is now a naked bet that can lose in full |
| The rulebooks differ | Extra time counted at one operator and not the other, or a player market that voids at one and loses at the other | Both legs can lose at once |
The first row is the common one. The leg you have not placed yet is exactly the one somebody else is also taking, for the reasons in why odds move before kick-off. Place the shorter, more volatile leg first.
The void row is the expensive one. On a ₦10,000 arb the legs are about ₦4,940 and ₦5,060; if the first is voided and the second loses, you are ₦5,060 down. At ₦370 a time that is 13.7 successful arbs erased by one refund. Which situations return a stake, and which do not, is set out in void bets and refunds — and the answer differs between operators, which is the whole problem.
One month, written out honestly
Assume ₦40,000 of working capital spread over four accounts, an average margin of 2.2% once the obvious opportunities have gone, and twenty-five arbs actually placed.
| Line | Amount |
|---|---|
| Turnover: 25 arbs × ₦6,000 | ₦150,000 |
| Gross at 2.2% | +₦3,300 |
| Five legs where the price moved before the second bet | −₦600 |
| Clean month, net | +₦2,700 |
| One voided leg whose pair lost | −₦3,000 |
| Month with one void, net | −₦300 |
The clean month returns 6.75% on ₦40,000 of capital — a decent figure until you divide it by the 35 or so hours of screen time twenty-five arbs take: about ₦77 an hour. The other month is negative, and nothing in it was done badly.
Important. Operator terms almost universally prohibit this. The practical consequence is not a legal one — it is voided bets, withheld bonuses, stake limits measured in hundreds of naira, and closed accounts. A restricted account rarely comes back, so the supply of usable accounts falls month by month while your capital stays the same.
Why it is a job rather than a strategy
Everything above describes an operation with fixed costs and a shrinking asset base: several funded accounts, prices to watch, cash sitting idle between settlements, and the discipline to place the second leg within seconds of the first. The reward per action is small; the failure modes are large and lumpy.
It is also self-limiting in a way that a method based on your own estimate is not: arbitrage burns accounts as fuel, and the better you are at it, the faster they burn. Priced as a part-time occupation, the hourly figure above is the whole argument.
Closing a position you already hold
The same arithmetic has a more common use: covering a bet you already placed, at prices that have since moved. You are not hunting a gap between operators, you are buying back one side of your own position.
Example. You hold ₦500 on Over 2.5 at 2.30, a potential return of ₦1,150. The match is 2-0 after twenty-five minutes and live Under 2.5 is available at 4.00. Stake the potential return divided by the opposite price: 1,150 ÷ 4.00 = ₦288. Total outlay is now ₦788. If a third goal arrives you collect ₦1,150 for a profit of ₦362; if it does not you collect 288 × 4.00 = ₦1,152, for ₦364. The result no longer matters.
Compare that with the one-button version. Cashing out the same position typically offers something below its fair value, and an offer of ₦800 locks ₦300 rather than ₦362 — the ₦62 difference is what you pay for convenience. The full comparison is in cash out explained.
Two conditions make it worth doing. The opposite price must be genuinely good, which means converting it first as in turning odds into probability. And the reason must be structural — a position too large for your bank — rather than nerves.
Where to look. Prices and live status for the current card sit on our match pages, which is where you would check a second price before covering anything.
FAQ
Is arbitrage betting legal?
Placing bets with two operators is not in itself unlawful, which is a different question from whether it is permitted. Operator terms almost always prohibit it, and enforcement is contractual: bets voided after the fact, stakes limited, accounts closed. Read the terms you accepted rather than assuming.
Can I run it with a small bank?
The arithmetic works at any size; the reward does not. A 3% arb on ₦500 pays about ₦15 for two apps, two logins and a wait for settlement. Below a few tens of thousands of naira in working capital, the hourly return is lower than almost any other use of the same time.
What happens if only one leg gets accepted?
You hold an ordinary single bet at whatever price you took. Decide in advance which you will do: take the closest cover even at a worse number, or let it run and accept the variance. The dangerous version is doing neither and leaving a half-hedged position unattended until kick-off.
This article is for information only and is not an inducement to gamble. Betting involves the risk of losing money — never stake more than you can afford to lose. 18+. If gambling stops being entertainment, read our responsible gambling guide and seek help.