The core concept · 6 min read
What is overround, and what is it costing you?
Every set of odds you have ever seen contained a fee. It is not hidden exactly, but it is never stated, and almost nobody calculates it. Here is how, in about ten seconds, with nothing but the prices in front of you.
The one calculation
Decimal odds contain a probability. A price of 2.00 implies a 50% chance, because 1 divided by 2.00 is 0.50. A price of 4.00 implies 25%. That is all a bookmaker's price is: an opinion about likelihood, with something added on top.
In a market where exactly one outcome must happen, the true probabilities have to add to exactly 100%. So take any market, convert every price, and add them up.
| Outcome | Price | 1 ÷ price | Implied chance |
|---|---|---|---|
| Home win | 2.10 | 1 ÷ 2.10 | 47.62% |
| Draw | 3.40 | 1 ÷ 3.40 | 29.41% |
| Away win | 3.60 | 1 ÷ 3.60 | 27.78% |
| Book total | 104.81% |
104.81% instead of 100%. That extra 4.81% is the overround, and it is the bookmaker's fee on that match. It is charged before a ball is kicked, it applies whichever result comes in, and it is the single largest determinant of how long a betting balance survives.
What the number should look like
Context matters, because more outcomes means a bigger total. On a three-way football match:
- Under 5%: sharp. You are getting close to a fair price.
- 5% to 7%: competitive. Normal for a well-priced book on a major league.
- 7% to 9%: expensive but common, especially on smaller competitions.
- Above 9%: you are paying a premium for convenience.
In our most recent South African measurement, the field ran from 6.00% at Playabets to 10.41% at Supabets. That is a spread of 4.41 percentage points on the same sport, in the same country, in the same week. See this week's full table →
The other way to say it: payout percentage
Divide 100 by the book total and you get the payout: 100 ÷ 1.0481 = 95.41%. That market returns about 95 cents in the rand across all outcomes over time. A 10% overround pays about 90.9 cents. Over a season those four and a half cents are the difference between a hobby and a slow bleed.
What it actually costs, in rands
A R200 bet through a 38-week football season, at a book charging 6% against one charging 10%. The fee is charged on every bet whether it wins or loses, so the only thing that changes the total is how often you bet.
| How often | Bets | Turnover | Cost at 6% | Cost at 10% | Difference |
|---|---|---|---|---|---|
| Twice a week | 76 | R15 200 | R860 | R1 382 | R521 |
| Five times a week | 190 | R38 000 | R2 151 | R3 455 | R1 304 |
| Every day | 266 | R53 200 | R3 011 | R4 836 | R1 825 |
Same bets, same matches, same results. The only difference is which app you opened. No prediction skill is required to keep that money, and it scales exactly with how much you bet.
Worked on the payout, not the headline margin: a 6% margin removes 5.66% of turnover, not 6%, because the margin is quoted against a base that already includes it.
Why racing numbers look worse
A fourteen-runner handicap has fourteen prices to hide a fee in. Racing markets routinely total 120% or more, and every individual price can still look reasonable. That is why comparing prices matters even more in racing than in football. More on racing markets →
The honest limit of this idea
A low overround does not make you a winner. It does not predict results and it does not turn a bad selection into a good one. What it does is stop you paying more than you need to for the same bet, every single time. That is the whole claim, and unlike almost everything else in betting, it is arithmetic rather than opinion.
Straight answers
Overround questions
What is overround in betting?
What is a good overround?
Is overround the same as the vig or the juice?
How do I calculate payout percentage from overround?
Does a low overround guarantee I win?
Now see it applied to real South African books
We run this exact calculation across every licensed bookmaker we can read, every Thursday morning, and publish the result.