The best way to use a bonus bet, if your aim is to keep as much of its value as you can, is usually on a longer price that sits close to its exchange price. A winning bonus bet pays only its winnings, so the stake it never returns costs least when the price is long. With illustrative exchange prices, a $50 bonus bet is worth about $23.81 at $2.00, $34.88 at $4.00 and $38.89 at $8.00 on average.
Longer prices also win less often, so the same choice makes the result swing harder, and the gap between the bookmaker's price and the exchange price matters as much as the length. The other way to use one, laying the selection on an exchange, gives up some of the average for a result that barely changes with the race, if the lay is matched at the price you planned.
The expected value of a bonus bet at any price
A stake-not-returned bonus bet pays bonus x (odds - 1) when it wins and nothing when it loses. Its value is therefore:
expected value = chance of winning x bonus x (odds - 1)
An exchange gives you the chance. On a betting exchange, punters set the prices by betting against each other, and the operator takes commission from net winnings rather than pricing in a margin. So 1 / exchange price can stand as an estimate of how often a selection wins. Put the two together and the share of the bonus you keep on average is:
share kept = (bookmaker price - 1) / exchange price
Two things follow. When the bookmaker's price equals the exchange price, the share is 1 - 1 / price: half the bonus at $2.00, three quarters at $4.00 and seven eighths at $8.00. And every cent the bookmaker's price sits below the exchange price comes straight off the top. Bonus bets explained covers why a winning bonus bet pays no stake.
Worked: a $50 bonus bet at $2.00, $4.00 and $8.00
Example: Illustrative prices, not real races or a current offer. You hold a $50 bonus bet, and runners in three different races fit its terms. Bookmaker A has them at $2.00, $4.00 and $8.00; the exchange has them at $2.10, $4.30 and $9.00. A fourth line shows the $8.00 runner again with an exchange price of $11.00.
| Bookmaker price | Exchange price | Chance it wins | A win pays | Expected value | Share of the $50 |
|---|---|---|---|---|---|
| $2.00 | $2.10 | 1 / 2.10 = 47.62% | $50 x 1.00 = $50.00 | $50.00 / 2.10 = $23.81 | 1.00 / 2.10 = 47.6% |
| $4.00 | $4.30 | 1 / 4.30 = 23.26% | $50 x 3.00 = $150.00 | $150.00 / 4.30 = $34.88 | 3.00 / 4.30 = 69.8% |
| $8.00 | $9.00 | 1 / 9.00 = 11.11% | $50 x 7.00 = $350.00 | $350.00 / 9.00 = $38.89 | 7.00 / 9.00 = 77.8% |
| $8.00 | $11.00 | 1 / 11.00 = 9.09% | $350.00 | $350.00 / 11.00 = $31.82 | 7.00 / 11.00 = 63.6% |
What the table shows:
- Moving from $2.00 to $8.00 lifts the average from $23.81 to $38.89. A winning $50 cash bet at $2.00 returns $100, half of it stake; at $8.00 it returns $400, an eighth of it stake. A winning bonus bet goes without that stake share every time.
- The gap to the exchange price matters as much as the length. The same $8.00 runner priced at $11.00 on the exchange keeps $31.82, less than the $4.00 runner at $34.88.
- These are averages over many bonus bets. Any single one pays its full win or nothing.
Risk: Betting involves risk. Every one of these bonus bets either pays its winnings or pays nothing, the exchange price is an estimate that moves, and there is no guarantee of profit. See responsible gambling for limits and support.
Longer prices win less often: the swings
The price that keeps the most value is also the one that pays nothing most often. Using the exchange chances from the table:
| Bonus bet at | Chance it pays nothing | Winners expected in 10 such bets | Chance all 10 pay nothing |
|---|---|---|---|
| $2.00 | 1 - 47.62% = 52.38% | 10 x 47.62% = 4.8 | 0.5238 to the power of 10 = 0.16% |
| $4.00 | 1 - 23.26% = 76.74% | 10 x 23.26% = 2.3 | 0.7674 to the power of 10 = 7.1% |
| $8.00 | 1 - 11.11% = 88.89% | 10 x 11.11% = 1.1 | 0.8889 to the power of 10 = 30.8% |
Ten bonus bets at $8.00 still average 10 x $350.00 / 9.00 = $388.89 between them, yet nearly one run of ten in three returns nothing at all. At $2.00 a run like that almost never happens, and the average is 10 x $23.81 = $238.10.
Where the terms let you split a bonus bet, smaller bets on separate races spread the result without changing the average. Five $10 bonus bets at $8.00 on different races still average $38.89 in total, and the chance that all five pay nothing is 0.8889 to the power of 5 = 55.5%, against 88.89% for one $50 bet.
A bonus bet puts none of your own money at risk, so a dry run costs no cash. It matters if you are counting on the bonus, and it matters most if a run of losing bonus bets tempts you to top them up with cash bets to get the money back.
Punt it, or convert it with an exchange lay
Converting means backing the selection with the bonus bet and laying the same selection on a betting exchange, so you keep about the same cash whichever way the race goes. This is matched betting applied to one bonus bet, and matched betting in Australia covers its lay maths in full, along with the ways a lay can go wrong. The table below only compares what a lay keeps with the punting average. The lay that balances the two sides is:
lay stake = bonus x (back odds - 1) / (lay odds - commission rate)
cash kept either way = lay stake x (1 - commission rate)
Example: The same three runners, with illustrative lay prices a little above the exchange back prices ($2.14, $4.40 and $9.20) and an illustrative commission rate of 6%, written 0.06. The rate is an input, not any exchange's actual rate.
| Back price | Lay price | Lay stake | Liability | Kept either way | Share of $50 | Punting average |
|---|---|---|---|---|---|---|
| $2.00 | $2.14 | $50.00 / (2.14 - 0.06) = $24.04 | $24.04 x 1.14 = $27.41 | $24.04 x 0.94 = $22.60 | 45.2% | $23.81 |
| $4.00 | $4.40 | $150.00 / (4.40 - 0.06) = $34.56 | $34.56 x 3.40 = $117.50 | $34.56 x 0.94 = $32.49 | 65.0% | $34.88 |
| $8.00 | $9.20 | $350.00 / (9.20 - 0.06) = $38.29 | $38.29 x 8.20 = $313.98 | $38.29 x 0.94 = $35.99 | 72.0% | $38.89 |
Longer prices convert better as well, but the lay ties up more money: $313.98 of liability to convert one $50 bonus bet at $8.00. Converting keeps less than the punting average, $38.89 - $35.99 = $2.90 less at $8.00 here, because commission and the gap between back and lay prices come off. What it buys is a near-equal result whichever way the race goes. With the lay stake rounded to the cent, the two outcomes can differ by a few cents: at $8.00 the runner winning leaves $350.00 - $313.98 = $36.02 and losing leaves $35.99, and the table shows what you keep if the runner loses.
The bonus bet converter does these sums for any prices. In racing, a late scratching can make the bookmaker bet and the lay settle differently, because the Stewards' deduction and the exchange's own adjustment need not match; matched betting on horse racing covers those cases.
Risk: Betting involves risk. A lay that is unmatched or only part matched leaves the bonus bet exposed, the liability has to sit in your exchange account, and prices can move between your two bets. See responsible gambling for limits and support.
What odds to use a bonus bet on: a decision rule
Run the same six checks each time:
- Read the terms: the expiry, any minimum odds, the markets and bet types allowed, and whether the bonus can be split.
- List the selections it could go on where you can see an exchange price.
- Work out (bookmaker price - 1) / exchange price for each, and favour the highest.
- If you plan to convert, also check that the lay price is close to the back price and that the liability fits your exchange balance.
- If you plan to punt, stay inside a price range whose swings you accept, and do not top the bonus up with cash.
- Check both prices again just before you bet.
Risk: Betting involves risk. Ranking by the ratio raises the average a bonus bet pays, not its chance of paying, and a run of long-priced bonus bets can return nothing many times in a row. See responsible gambling for limits and support.
Racing fields offer long prices, usually with an exchange market to check them against. A two-way sports market near $1.90 keeps at most 1 - 1 / 1.90 = 47.4% even at a fair price, so a bonus bet there behaves like the $2.00 row. Multis reach long prices quickly but carry a margin on every leg and cannot be laid as one bet: bonus bets on multis works through them.
Mistakes that waste a bonus bet
Each mistake, costed with this page's illustrative prices and two more illustrative ones, a $1.60 favourite against an exchange $1.65 and a lay at $10.00:
| Mistake | What it costs |
|---|---|
| A short favourite: $1.60 against an exchange price of $1.65 | $50 x 0.60 / 1.65 = $18.18 on average, $16.70 less than the $4.00 runner |
| The longest price, whatever the gap | $8.00 against $11.00 averages $31.82, $7.07 less than against $9.00 |
| A lay at a wide price | Laying the $8.00 runner at $10.00 instead of $9.20 keeps $350.00 / 9.94 x 0.94 = $33.10, about $2.90 less |
| Letting it expire | The whole average: $34.88 on the $4.00 runner |
| Topping it up with cash bets | Each cash bet carries the bookmaker's margin, so every extra dollar bet loses a little on average |
Skip the bonus bet altogether if the only selections it fits are ones you would not otherwise back and you feel the pull to add cash. Letting it lapse costs no cash, and betting promotions in Australia sets out how to opt out of offers that push you to bet more.
Bonus settings in B337's automation
In B337, bonus betting belongs to the Classic Racing strategy, which runs under Full Automation and works only with bonus bets a bookmaker has already put on your own accounts. It bets a bonus only on a runner within your odds range whose conversion, measured against the Betfair price of the moment, reaches your threshold. If no runner passes, it places nothing, and the bonus waits on the account until it is used or runs out under the bookmaker's terms. B337 never creates or claims a bonus bet.
The strategy's bonus settings, from the odds range and conversion threshold to a bonus max stake and optional stacking, are set out on bonus bet automation.
Only the bookmaker side is automated. B337 never places the lay, so any bonus bet a session puts on stays unhedged until you hedge it yourself in your own exchange account. The bot lives on your own computer and places nothing while that machine is off, asleep or offline. Full Automation is set up with the team, which confirms the price, including any charges on top of the plan price, before you pay; bets placed through B337 use credits, a per-bet usage charge. Bookmaker names are trade marks of their owners. B337 is not affiliated with them.
For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au.
Risk: Betting involves risk. Many bookmakers restrict automated betting in their terms, and any of them may reduce your stakes, void bets or shut the account. Automation can bet far more, far faster, than by hand, so a deposit limit with each bookmaker limits how much new money it can risk. A bonus bet that loses pays nothing. See responsible gambling for limits and support.