Bookmaker Odds Disagreement — How Far Apart Are Australian Prices?
Same outcome, same moment, different price. Across 18,799 head-to-head outcomes the typical bookmaker disagreement is 2%–7% — small, consistent, and worth more than most bettors think.
Updated · 7 min read
Key takeaways
- Measured 18,799 priced head-to-head outcomes across 8,692 event slots and 41 sports in the 30 days to 30 October 2026.
- Typical disagreement — the gap between the best and worst price at three or more books — is 2.22% to 6.67% for the mainstream sports.
- Soccer is the widest of the big four at a 6.67% median spread; tennis is the narrowest at 4.00%, with basketball at 3.28% and baseball at 2.89%.
- The widest sport in the raw data, ice hockey at a 27.04% median, is three books priced 21%–31% above an otherwise identical consensus — not a market-wide disagreement.
- Price shopping pays most where the median spread is largest: on a $100 bet a 4% gap is $4, and soccer's p90 spread is 22.92%.
What we measured
For every event in the archive we took the head-to-head market and collected one price per bookmaker for each outcome. An outcome only entered the sample if at least three books quoted it, so a two-book comparison could never masquerade as a market. Over the 30 days to 30 October 2026 that left 18,799 priced outcomes across 8,692 event slots and 41 sports.
Two numbers describe each outcome. The spread is the gap between the best and worst price as a percentage of the worst — how much you gain by shopping. The coefficient of variation is the same idea using every book's price, so one wild outlier moves it less. We report the median and the 90th percentile of the spread within each sport, because the median tells you what a normal bet looks like and the p90 tells you what is worth hunting for.
The headline table
Clean sample: prices under 20.00, at least three books per outcome, 30 days to 30 October 2026.
Median and 90th-percentile spread by sport
- tennis · 5,997 outcomes · 3,069 events · median 4.00% · p90 14.49%
- soccer · 3,329 outcomes · 1,233 events · median 6.67% · p90 22.92%
- basketball · 1,349 outcomes · 711 events · median 3.28% · p90 17.97%
- darts · 902 outcomes · 461 events · median 6.00% · p90 32.08%
- ice hockey (NHL) · 616 outcomes · 249 events · median 27.04% · p90 56.16%
- ice hockey (other) · 516 outcomes · 281 events · median 23.53% · p90 44.23%
- baseball · 388 outcomes · 196 events · median 2.89% · p90 8.09%
- American football · 374 outcomes · 188 events · median 2.34% · p90 10.00%
- cricket · 288 outcomes · 143 events · median 4.78% · p90 15.08%
- snooker · 203 outcomes · 102 events · median 6.67% · p90 58.67%
- volleyball · 175 outcomes · 91 events · median 2.22% · p90 12.50%
- rugby union · 172 outcomes · 90 events · median 5.15% · p90 22.14%
- MMA · 116 outcomes · 60 events · median 3.36% · p90 11.36%
The ordering is stable: the sports with the deepest liquidity and the most mature pricing — tennis, basketball, baseball, American football, volleyball — cluster at the narrow end, while soccer, darts and snooker sit wider. A 4% median spread on tennis sounds trivial until you apply it to a season of bets: taking the best price rather than the worst on every tennis wager is worth roughly 4% of turnover before commission, which is more than most staking plans generate.
The p90 column is where the value is. On soccer, one outcome in ten is priced 22.92% apart across books. Those are the bets where a comparison is the difference between a marginal edge and no edge at all.
The widest sport is an artefact, not an edge
Ice hockey shows a 27.04% median spread, more than six times soccer. It is not a market where books genuinely disagree about hockey. Breaking it down by book, three of them — tabtouch at 1.308x the consensus price, unibet at 1.242x and betright at 1.212x — sit consistently 21% to 31% above everyone else, while every other book lands between 1.000 and 1.041.
In other words, the "widest disagreement in the archive" is a handful of books quoting a different number from an otherwise unanimous market. That is worth knowing for two reasons: it is a comparison opportunity if those prices are real and takeable, and it is a warning that a large measured spread is more often a single source priced away from the field than a market in flux. We have kept the sport in the table rather than dropping it, because the pattern is the finding.
What it means for a bet
Price shopping is the only edge available to every bettor on every bet. It requires no model, no timing and no market knowledge — just the discipline to check two or three books before staking. This dataset sizes the prize: on mainstream sports it is worth 2%–7% of stake on a typical bet and considerably more in the tail.
It also sets an honest ceiling. A 6.67% median spread on soccer is smaller than the margin most bookmakers build into a soccer head-to-head market, so shopping alone will not turn a losing bettor into a winning one. It converts a losing bet into a less-losing one, and on a bet that was already marginal it is often the whole margin.
The practical rule the data supports: the less liquid and less mainstream the sport, the more the price varies and the more a comparison matters — but the wider the bookmaker margin you have to overcome to realise it.
Method and limits
Sample: 18,799 head-to-head outcomes with three or more quoting bookmakers, drawn from 8,692 event slots across 41 sports in the 30 days to 30 October 2026. Prices above 20.00 were excluded from the headline cut because a single long-priced error inflates a percentage spread without describing the market; the unfiltered distribution is dominated by that tail.
Two cuts were run. The clean cut excludes prices at or above 20.00. The fresh cut keeps only books whose price timestamp sits within 15 minutes of the newest book on the same event, so that a pre-match price is never compared against an in-play one. Both cuts produce the same sport ordering, which is why the headline figures are quoted from the clean cut.
Limits: multi-sport events appear under more than one competition key, so event counts are not strictly disjoint. Sport names are our own canonical keys and include per-competition entries such as soccer's domestic divisions. Percentages are medians of per-outcome spreads, so a sport with many lopsided outcomes contributes many small spreads; read the p90 alongside the median. Figures are as at 30 October 2026 and are recomputed from the archive, not hand-picked.
FAQ
How much do Australian bookmakers disagree on the same odds?
For mainstream sports, the typical gap between the best and worst price on the same head-to-head outcome is 2.22% to 6.67% of the worst price. Soccer is widest of the big four at a 6.67% median, tennis narrowest at 4.00%, with basketball at 3.28% and baseball at 2.89%.
Which sport has the biggest price differences between bookmakers?
In the raw data ice hockey shows a 27.04% median spread, but the cause is three books quoting 21%–31% above an otherwise identical consensus rather than genuine market disagreement. Among sports where books are broadly aligned, soccer (6.67%) and darts (6.00%) sit at the wide end.
Why do bookmakers disagree on the same price at all?
Different books weight their own customer flow, their risk exposure and different sharp reference markets. Most mainstream books converge to within a few percent, which is why the median spread is small and the tail — where one book has not yet moved — is where the value sits.
Does a wide price gap mean one bookmaker has made a mistake?
Sometimes, but often it means one book has updated and another has not, or that the two prices are not from the same moment. We only compare prices stamped within 15 minutes of each other for the fresh cut, and we exclude prices of 20.00 and above from the headline figures, because those inflate a percentage spread without describing a real market.