How Bookmakers Make Money: Margin, Odds and the Built-in Edge
A bookmaker sells every outcome at slightly worse than its fair price. Follow one match from an estimated probability to the odds on your slip, and see what the operator keeps whoever wins.
A bookmaker does not make money by predicting results. It makes money by selling every outcome of a match at slightly worse than its fair price, then keeping a share of everything staked whichever way the game goes — about five kobo in the naira on a normal three-way football line. This page follows one match from an estimated probability to the price on your slip, and explains why a punter who picks more winners than losers can still finish the season behind.
In short.
- The business model is turnover multiplied by a retained percentage, not clever forecasting.
- A fair 52% chance is worth 1.92. Priced up by roughly 5%, it reaches your slip at 1.82.
- On a perfectly balanced book of ₦100,000 the operator pays out about ₦94,700 whichever outcome wins, keeping ₦5,300.
- Prices move to pull money towards the light side of the book, not because someone changed their mind about the match.
- At 1.82 you need to be right 54.9% of the time simply to break even — being right more often than you are wrong is not the same as winning.
The business is a percentage of turnover
Think of a bookmaker as a shop with one product: risk. Punters bring money, the shop hands back a fixed proportion of it and keeps the rest. That proportion is set in advance by the prices, which is why an operator can forecast revenue without knowing a single result.
Three mechanisms protect it: the mark-up built into the odds, the price movements that stop any one result mattering too much, and the limits applied to customers who beat it. Those three explain your own chances better than any tipping service.
From an estimate to a price on your slip
Pricing starts with a probability. A model, adjusted by a trader, produces something like 52% home, 26% draw, 22% away. Those add to 100%, so the fair odds are one divided by each figure — the conversion in how to turn odds into probability, run backwards.
Then the mark-up goes on. Multiply every probability by about 1.055, convert again, and the fair prices shrink into the ones you see.
| Outcome | Estimated chance | Fair price | Price offered | What the shortening costs you |
|---|---|---|---|---|
| Home win | 52% | 1.92 | 1.82 | ₦50 less per ₦500 staked |
| Draw | 26% | 3.85 | 3.65 | ₦100 less per ₦500 staked |
| Away win | 22% | 4.55 | 4.30 | ₦125 less per ₦500 staked |
Read the offered column back as probabilities — 54.9%, 27.4%, 23.3% — and they total 105.6%. That excess is the whole business. As a share of money staked it is 1 − 1 ÷ 1.056 = 5.3%: about 5.3 kobo of every naira passing through this market stays behind.
Two-way markets cost less than three-way ones
The retained share is not a constant. It depends on how many outcomes are priced and how hard operators compete for the bet. The formula never changes: add up the implied probabilities, then compute 1 − 1 ÷ total.
| Market | Prices | Implied total | Retained per ₦1 | Cost of ₦12,000 turnover |
|---|---|---|---|---|
| Two-way, tight | 1.95 / 1.95 | 102.6% | 2.5 kobo | ₦300 |
| Two-way, typical | 1.87 / 1.93 | 105.3% | 5.0 kobo | ₦600 |
| Three-way | 1.82 / 3.65 / 4.30 | 105.6% | 5.3 kobo | ₦636 |
| Three-way, wide | 1.75 / 3.60 / 4.10 | 109.4% | 8.6 kobo | ₦1,032 |
The last column translates those percentages for an ordinary punter. Put a ₦12,000 monthly budget through the tight two-way market once and the built-in cost is ₦300; through the wide three-way market it is ₦1,032. Nothing else about your betting changed — not your selections, not your luck.
Extra outcomes also give the trader more places to hide the mark-up. On a two-way line a shortening is obvious, because both prices sit side by side. On a three-way line the draw and the outsider can carry more of the load, and few punters add the three numbers up.
Balancing the book with other people's money
The ideal is a book where the money on each outcome is proportional to its price, so the payout is identical whoever wins. Take ₦100,000 of turnover on the line above.
| Outcome | Share of turnover | Staked | Paid out if it wins | Operator keeps |
|---|---|---|---|---|
| Home at 1.82 | 52.0% | ₦52,030 | ₦94,695 | ₦5,305 |
| Draw at 3.65 | 25.9% | ₦25,940 | ₦94,681 | ₦5,319 |
| Away at 4.30 | 22.0% | ₦22,030 | ₦94,729 | ₦5,271 |
The last column is almost identical in all three rows: the match has become irrelevant and about ₦5,300 stays behind — the 5.3% worked out earlier. This is the state the trading desk aims at, and it is why prices move.
When the money lands on one side
Real books rarely balance by themselves. Public money follows famous teams, so a Premier League favourite at home attracts most of the stakes unaided. Suppose ₦70,000 of the ₦100,000 goes on the home side, ₦18,000 on the draw and ₦12,000 away.
Example. Home wins: ₦70,000 × 1.82 = ₦127,400 paid out against ₦100,000 taken — a loss of ₦27,400. Draw: ₦18,000 × 3.65 = ₦65,700, a profit of ₦34,300. Away win: ₦12,000 × 4.30 = ₦51,600, a profit of ₦48,400.
Weight those results by the estimated probabilities: 0.52 × (−₦27,400) + 0.26 × ₦34,300 + 0.22 × ₦48,400 = +₦5,318 — the same margin as the balanced book. The expectation did not change; the variance did. One result now costs ₦27,400 instead of nothing.
So the home price is cut towards 1.75 and the away price pushed out to 4.60: a shorter favourite attracts less money, a longer outsider more, and the stakes drift back towards the balanced split. The move is risk management, not a prediction — a distinction covered from the punter's side in betting odds explained.
Why picking winners is not enough
The mark-up sets a pass mark you have to clear before any of your skill counts. At a price of 1.82 the break-even hit rate is 1 ÷ 1.82 = 54.9%. Win 54% of your bets and you lose money; win 56% and you make some.
| Price taken | Break-even hit rate | Hit rate needed for 5% on turnover |
|---|---|---|
| 1.50 | 66.7% | 70.0% |
| 1.82 | 54.9% | 57.7% |
| 2.30 | 43.5% | 45.7% |
| 3.65 | 27.4% | 28.8% |
| 4.30 | 23.3% | 24.4% |
Put numbers on it. A hundred ₦500 bets at 1.82 with 55 winners return 55 × ₦910 = ₦50,050 against ₦50,000 staked: ₦50 of profit for a season of beating the market. Fifty-four winners return ₦49,140 — a loss of ₦860 on the same selections.
The gap between 54 and 56 winners is the entire business — and the reason the honest test of a bettor is the price taken rather than the count of correct calls. That is why reading a price properly matters more than picking a side.
Limits: what happens to accounts that win
Because the edge is thin, operators protect it by managing customers rather than matches. Accounts are profiled on behaviour, and the patterns that attract attention are the ones a profitable bettor produces naturally: taking prices minutes after they open, staking odd amounts calculated from a bankroll rule, betting into thin markets and beating the price the market closes at.
Restriction is rarely announced. The usual signs are a maximum stake falling from thousands of naira to a few hundred, bets held for manual approval, and promotions quietly disappearing from the account.
Important. A limit is a compliment with a cost. It confirms the account has been profitable, and it caps how profitable it can ever be — which is why serious bettors treat limits as a planning problem rather than a grievance.
The consequences are modest but real. Keep accounts with more than one operator, size stakes by rule rather than by confidence — see bankroll management — and accept that any strategy needing large stakes to be worth the effort has a short shelf life.
FAQ
Do bookmakers lose money on some matches?
Regularly. A book weighted towards the favourite loses when the favourite wins, as in the ₦27,400 example above. What protects the operator is repetition: thousands of markets a week, each priced with a few percent in hand, so individual losses average out the same way individual punters' wins do.
Is the margin the same on every market?
No. It varies with the number of outcomes and with competition for the bet. In the table above the retained share ranged from 2.5% on a tight two-way line to 8.6% on a wide three-way one — a difference of ₦732 on a ₦12,000 monthly turnover, before any selection skill is involved.
Why did my stake get reduced to a few hundred naira?
The account has most likely been flagged as profitable or as showing patterns associated with profitable customers. It is a commercial decision made by the operator, not a penalty, and it usually appears as a smaller maximum stake or as manual approval on bets rather than as a formal notice.
If the margin is only 5%, why do most punters lose much more?
Because the 5% is charged on every bet, not once. Stake ₦12,000 and recycle the returns four times over a month and you have put roughly ₦40,000 through a market that keeps five kobo in the naira. Accumulators compound the effect further, since each leg carries its own mark-up.
Read next
- Betting odds explained — formats, implied probability and how to spot a price worth taking.
- The match centre — compare the current prices on today's fixtures before you stake.
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.
