For offers that only pay out on one result: stake back as a free bet if your bet loses, a bonus if it wins. Set the lay stake here and both results pay you the same, whichever way it goes.
Lay less. All of the profit lands if the bookmaker bet wins.
Lay more. All of the profit lands on the side that pays the bonus.
Some bookmaker offers only trigger on one result. Bet £20, and if the selection goes down you get your stake back as a free bet. If it wins, you win, and that's the end of it. A few offers run the other way and pay the bonus when the bet wins.
Back and lay one of those in the ordinary way and you finish with two different answers: a few pence either side if the bet wins, and a free bet sitting in your account if it loses. This page picks the single lay stake that makes both results pay the same figure. You know the number before the game kicks off, and nothing about the score changes it.
The move is to lay a little less than a normal qualifying bet would ask for. That leaves you slightly short if the selection wins, and the refund covers exactly that gap when it loses.
For the common version, refund as a free bet if the selection loses:
Lay stake = (back stake × (back odds − 1) × (1 − back commission) + back stake − bonus value) ÷ (lay odds − lay commission)
Take away the middle term and you have the standard qualifying lay stake. The only change here is subtracting the bonus value, which is the bonus multiplied by the share of it you expect to turn into cash. Lay commission is a fraction, so 2% goes in as 0.02.
If the bonus lands on a winning bet instead, the bonus value moves to the other side of the sum and is added rather than subtracted. The calculator flips it for you when you change the switch.
A £20 bet, stake back as a free bet if it loses. The oddsmatcher finds a selection at 5.00 with the exchange laying 5.20, commission 2%, and we'll assume 80% of the free bet converts, so the bonus is treated as £16.
Same figure either way, decided before kick-off. The difference is only in how you get there: one result pays you in cash on the day, the other pays you in cash plus a free bet you still have to convert.
The second bet is where the 80% assumption gets tested. Lock in £11.89 on the basis of an £16 free bet, convert it for £14, and you're £2 up on the plan. Convert it for £13 and you're £3 down on it. The lock-in is only as honest as the retention figure you feed it.
Standard is the default and it's the right answer almost every time. It splits the difference so both results pay the same, which is the point of the exercise.
The other two shift the money to one side on purpose.
Anywhere between the two lay stakes, both results are in profit. That band is the calculator's own sanity check: if the underlay figure comes out above the overlay figure, no lay stake makes both sides positive and the offer doesn't clear at those prices. Better odds or a better price will fix it. Forcing it will not.
Longer odds put more of the bonus in your pocket, because the refund triggers more often when the selection is less likely to win. The same £20 offer, £16 of assumed bonus value, 2% commission, at three different prices:
| Back | Lay | Lay stake | Liability | Locked in |
|---|---|---|---|---|
| 2.00 | 2.06 | £11.76 | £12.47 | £7.53 |
| 5.00 | 5.20 | £16.22 | £68.11 | £11.89 |
| 10.00 | 10.50 | £17.56 | £166.79 | £13.21 |
Going from evens to 5.00 adds £4.36 for nothing but a different selection. Going from 5.00 to 10.00 adds £1.32 and asks for another £99 of liability, which is a poor trade unless your exchange balance is deep and idle. Four to eight is the useful range on most refund offers. Above that you're renting out a large chunk of your bank to chase pennies, and the gap between back and lay usually widens too, which quietly eats the difference.
Retention is the share of the free bet you expect to end up with in cash. It's the one input on this page that is an estimate rather than a fact, and it moves the answer.
When you aren't sure, go low. An offer that still looks fine at 70% is an offer that will look fine on the day. One that only works at 85% is an offer built on a number you have to hit perfectly, and there's no reason to take that on.
Because the refund only turns up on the losing side, so that side needs less help from the exchange. A standard qualifying lay balances the two results on their own merits. Here the bonus is already sitting on one of them, and the lay stake shrinks by roughly the value of that bonus divided by the lay odds.
Around 75% to 80% for an ordinary stake-not-returned free bet with no minimum odds attached. Drop it to 70% if the bookmaker forces you into odds of 2.00 or higher, or if the free bet is split into small chunks with an expiry you might miss. Push it to 100% only when the refund lands as withdrawable cash.
No, that is the offer working as designed. On a refund-if-it-loses offer a winning bet means no bonus, and the profit on that side is already in your account from the back bet against the lay. You are finished with the offer and the figure should match what the calculator showed for the bookmaker-wins result.
Laying at long odds means covering the full winnings if the selection comes in, so a £17 lay at 10.50 ties up about £160 in the exchange until the market settles. That money is not at risk in the sense of being lost, but it has to be sitting there. It is the practical ceiling on how long the odds you pick can be.
Yes, set the retention to 100% and the sums handle it. Cash back is the better version of the same offer because there is no second bet to place and nothing to convert, so the figure the calculator locks in is the figure you keep.
Every refund offer we've priced sits on the betting offers page, and we post them in the Discord as they land if you'd rather not keep checking back.
18+. Only stake what you can afford to lose. BeGambleAware.org.
Record your profit and expected value