On 12 October 2026 we read every bookmaker's win prices on 50 Australian greyhound races and computed the margin in each one. That is 402 bookmaker-races across the thirteen bookmakers we name publicly, and the answer is that greyhound racing is the most expensive thing on an Australian betting board by a wide margin.
The companion study, Australian bookmaker margins measured, ran the same arithmetic over 5,245 sports markets on the same day and found the same bookmakers charging between 6.95% and 10.11%. On greyhounds they charged between 27.32% and 39.15%.
How a win-market margin is calculated
Turn each runner's price into an implied chance, which is 1 divided by the decimal price. Add every runner in the field. A market with no margin sums to 1.00, and the excess is the bookmaker's share.
Example: an eight-dog field priced at $6.00 across the board implies 16.67% per runner, which sums to 133.33%. The overround is 33.33%, and on a $100 bet that is an expected cost of $25.00.
The field size is why greyhound numbers look so large next to sports numbers. The same cost per dollar staked, spread across eight runners instead of two, produces a much bigger overround figure while feeling less obvious on any single price.
What was measured, and what was thrown away
50 races, 402 usable bookmaker-races, from the Australian meetings open at the time of sampling. Three exclusions do real work, and each one changes the answer if you skip it.
Scratched runners: 70 excluded. A scratching leaves a dead price on the card. Add it to the sum and you inflate the margin by that runner's entire implied chance, which on a short-priced scratching is enormous. This is the single easiest way to publish a wrong number here.
Part-priced fields. A bookmaker that prices nine runners in an eleven-dog field has no measurable overround: the sum is short because prices are missing, not because the book is generous. Those are skipped, and the skip counts are published in the table below because they are large.
Stale prices: 8,199 skipped. Only prices changed within 45 minutes of sampling were used. That number is bigger than the number we kept, which is itself worth knowing: most books only move a greyhound price as the race approaches.
Margin by bookmaker
Median win-market overround, greyhound races only, 12 October 2026.
| Bookmaker | Median margin | Races measured | Part-priced skipped | Cost per $100 |
|---|---|---|---|---|
| BetRight | 27.32% | 37 | 5 | $21.46 |
| Sportsbet | 29.77% | 31 | 15 | $22.94 |
| TABtouch | 31.13% | 29 | 19 | $23.74 |
| bet365 | 32.03% | 35 | 13 | $24.26 |
| Palmerbet | 32.25% | 31 | 12 | $24.39 |
| Ladbrokes | 33.06% | 29 | 17 | $24.85 |
| Betr | 33.56% | 31 | 13 | $25.13 |
| Neds | 33.61% | 29 | 17 | $25.16 |
| TAB | 33.89% | 39 | 6 | $25.31 |
| PointsBet | 34.09% | 27 | 18 | $25.42 |
| Unibet | 34.61% | 24 | 14 | $25.71 |
| Picklebet | 37.51% | 32 | 9 | $27.28 |
| Dabble | 39.15% | 28 | 14 | $28.14 |
The spread is 11.83 percentage points, or $6.68 per $100 staked between the narrowest and the widest book. That is a far bigger gap than the 0.89 points separating eleven of the same bookmakers on sports, and it is the clearest practical reason to compare prices on greyhounds specifically.
The skipped column deserves attention rather than a footnote. A bookmaker that frequently prices only part of a field is offering you less of the market than the race-count suggests, and that is a coverage fact the margin figure on its own does not carry.
When you look changes what you see
Every observation was bucketed by how long remained before the jump. Across all books measured:
| Time before the jump | Median margin | Observations |
|---|---|---|
| More than 3 hours | 36.30% | 119 |
| 1 to 3 hours | 34.85% | 1,683 |
| 15 to 60 minutes | 34.76% | 1,411 |
| Inside 15 minutes | 28.70% | 1,270 |
The book tightens by about 7.6 points over the life of the market, and almost all of that happens in the final quarter hour. A margin figure quoted without the sampling window attached is close to meaningless, which is why this one carries it.
Risk: Betting involves risk. Bookmakers can restrict or close accounts and void bets, prices move, and a positive expected value bet can still lose. A narrower margin is not a winning bet and there is no guarantee of profit. See responsible gambling for limits and support.
Field size
| Runners priced | Median margin | Races |
|---|---|---|
| 2 to 7 | 32.11% | 297 |
| 8 to 11 | 36.28% | 112 |
A full field costs more than a depleted one, which follows from the arithmetic: more runners, more prices, more places to hold margin. A race that has lost runners to scratchings is a cheaper race to bet, before anything about the dogs is considered.
Two brands, nearly one price
The sports study found Ladbrokes and Neds quoting identical prices on all 1,049 selections compared. On greyhounds the same pair agreed on 173 of 174 runners, which is 99.43%.
The honest caveat is that the baseline is different here. Greyhound win prices move in coarse steps and unrelated bookmakers agree far more often by chance than they do on sports: among the other named pairs we measured, agreement ran from 13.3% to 86.25%. So 99.43% is still clearly apart from the field, but it is not the same stark result as 100% against a 5% baseline, and presenting it that way would overstate it.
We also tried to count how many of the 159 brands in the data are genuinely distinct price sources. That number moved between 47 and 63 across two runs of the same method, because a small brand needs enough complete fields before it can be compared at all. An unstable number is not a finding, so it is not published here. A larger sample would settle it.
What this is not
It is not thoroughbred or harness racing. When this sample was taken, on a Sunday evening, every Australian thoroughbred race for the day had already closed and only four harness races remained open. Nothing above describes those codes, and anyone extending it to them is guessing.
It is not the tote. A tote dividend does not exist until the pool closes, so there is no price to sum, and the takeout works differently. It is not the exchange either, where the cost is a commission on winnings rather than an overround.
And it is not a ranking of who gives the best price on your dog. Which bookmaker has the best odds is a different question, answered one selection at a time rather than one field at a time.
The prices came from the same board shown on odds comparison, read on a repeating cycle rather than tick by tick, so any individual price can trail the bookmaker's own by the time it is read.