How cash out is calculated in football betting, and what it costs
Published
Understanding how cash out is calculated in football betting takes one formula and one uncomfortable observation. The formula is simple enough to do on a phone: your potential return, divided by the odds at which the operator will let you bet against yourself. The observation is that those odds are never the fair ones. Every cash-out offer is a hedge executed at a price you did not shop for, and the difference between the offer and the fair value of your ticket is a fee — a second fee, on a position you already paid a margin to open.
How cash out is calculated in football betting: the formula
Take a hypothetical bet: £20 on the home win at decimal 3.40. Potential return is £68.00, of which £48.00 is profit.
An hour later the match state has changed and so has the probability of that home win. Suppose the fair probability is now 55%. The fair value of your ticket is what it would be worth to someone pricing it honestly:
0.55 × £68.00 = £37.40
That is the same thing as dividing the return by the fair odds — 1 ÷ 0.55 gives 1.818, and £68.00 ÷ 1.818 is £37.40. Both routes land in the same place, because a ticket is worth its payout times its chance of paying.
Now the operator quotes you a number. It does not plug 1.818 into that division. It plugs in something longer — say 1.95:
£68.00 ÷ 1.95 = £34.87
You have been offered 93.2% of what the ticket is worth. The missing £2.53 is not an error and it is not a rounding artifact. It is the price of the button.
Where the second margin hides
The odds used in a cash-out calculation are not the odds on the board next to the home win. That price sits on the short side of fair — an operator quoting a 55% chance will offer you something nearer 1.75 to back it, because it is selling. Cash out is the opposite trade: the operator is buying your position back, and it buys on the long side of fair for exactly the same reason. The margin flips direction but never disappears.
This is why cash-out offers hold a roughly constant share of fair value regardless of whether you are winning or losing. The same hypothetical £20 at 3.40, at three different moments:
| Match state | Fair probability | Fair value | Cash-out odds used | Offer | Fee |
|---|---|---|---|---|---|
| 0–0, half-time | 55% | £37.40 | 1.95 | £34.87 | 6.8% |
| 1–0 home, 60 minutes | 78% | £53.04 | 1.37 | £49.64 | 6.4% |
| 0–1 home, 60 minutes | 7% | £4.76 | 15.30 | £4.44 | 6.7% |
The fee is proportional. Cashing out to rescue something from a losing position costs the same percentage as cashing out to bank a winner — it just looks smaller because the number it is taken from is smaller.
The more important point is that this is the second time you have paid. If the fair price at kick-off was 3.60 and you took 3.40, you already handed over part of your edge on the way in. Converting decimal odds to implied probabilities makes that first charge visible: 1 ÷ 3.60 is 27.8%, 1 ÷ 3.40 is 29.4%, and the gap is the operator’s cut. Our guide to converting odds to implied probability walks through how to strip that out of a full board. Cash out charges the same kind of fee again, on the same position, in the opposite direction.
The hedge you could have placed yourself
Cash out is not a special product. It is a hedge, and you can price the same hedge by hand.
Back at half-time, with the home win fairly priced at 55%, the other side is fairly priced at 45% — decimal 2.222. To lock the position you would stake enough on “not home” that both outcomes pay the same:
£68.00 ÷ 2.222 = £30.60
Stake that, and whatever happens you collect £68.00 against a total outlay of £50.60. Locked profit: £17.40, which on your original £20 is equivalent to a cash-out value of exactly £37.40 — the fair number, as it must be.
Now do it at a real price. If the best available price on “not home” is 2.10 rather than 2.222, you need £32.38 to level the position, your profit falls to £15.62, and the equivalent cash-out value is £35.62. Still worse than fair. Still better than the £34.87 the button offered.
That £0.75 gap is the whole argument in miniature. Manual hedging is clumsy and lets you shop the opposing price; the button is instant and lets you shop nothing. Neither is free, and the convenient one costs more.
Two things make the gap wider rather than narrower. On an accumulator, the operator prices every remaining leg to buy back, so the fee compounds across legs instead of applying once. And a partial cash out charges the identical percentage on the portion you close, so taking half out twice costs the same as taking all of it once.
What cashing out does to a record
There is a reason this matters beyond the arithmetic, and it has to do with what a betting record is supposed to prove.
A prediction is a claim about a match. It is settled by the match. The moment you cash out, the thing that settles your bet is no longer the result — it is a decision you made in the sixty-first minute, using information the prediction never had. A record that mixes cashed-out bets with settled ones is not measuring the quality of anyone’s forecasts. It is measuring a forecast plus an unrecorded series of in-play judgement calls, and the two cannot be separated afterwards.
That ambiguity is convenient. Cash out is the easiest way to make a record look smooth: close the winners early for a slightly reduced return, close the losers early for a small recovery, and a volatile ledger turns into a tidy one that never quite shows a full-priced loss. Nothing about it is dishonest on any single bet. In aggregate, it makes the record unreadable, and no amount of auditing after the fact can put back the information the early exit destroyed.
This is why every prediction we publish is fixed before kick-off and settled at full time, at the published price, win or lose. Our methodology states when a prediction is made and what settles it; the track record page carries all of them on the same terms. There is no early-exit column, because an early exit is not a prediction outcome.
When it is still reasonable
None of this makes cashing out irrational. It makes it a purchase.
In pure expectation, closing a position early is negative: you are paying a few per cent of fair value to remove uncertainty. The defensible reasons to do that are about something other than expectation. A stake that is genuinely too large for your bankroll is worth de-risking even at a fee — that is a staking error being corrected, not an edge being harvested. And if you have watched the match and hold information the pre-match price genuinely lacks, you may be trading at odds you believe are wrong in your favour, which is a real reason even if it is hard to verify about yourself.
What does not work is the habit that cash out is designed to encourage: taking a small profit early, routinely, because a locked-in number feels better than an open one. Do that across a season and you have converted a thin edge into a reliable stream of fees, paid to close positions that would mostly have settled fine on their own. The arithmetic is unsentimental about it. The formula runs the same way every time, and the operator is always on the long side of fair.