Horse racing odds.
Five bookmakers compared.

BetRight led this sample

BetRight offered the highest or joint-highest win price on 42 of 71 runners. Sportsbet was next on 35. We compared the same runners across Sportsbet, Betr, BetRight, PlayUp and Unibet.

71 matched runners · all five bookmakers priced every runner · a tied highest price counts for each bookmaker.

The split matters. BetRight stood alone at the highest price on 26 runners; Sportsbet did so on 11 and Betr on four. PlayUp and Unibet shared the lead on some runners but never held it alone.

The median price gap was 25%

For each horse, we divided the highest decimal price by the lowest, then subtracted one. The median difference was 25%. Two runners had identical prices across all five sources.

The gaps were wider at longer prices. Horses with median odds of $15 or more had a 53.3% median gap, against 5.7% for horses below $3. Every runner belongs to one price group.

Median percentage difference between the highest and lowest decimal win price. Groups use the median of all five quotes for each horse.

A real runner, all five prices

Last Ditch Effort in Mount Isa R4 was priced from $12.00 to $15.00. That is a 25.0% difference in quoted decimal odds. We chose this runner because its gap was closest to the sample median.

A higher quoted price changes the potential payout on the same stake. It does not establish a positive expected return.

One good price isn’t the whole market

Overround uses every active runner in a race. We added each runner’s implied probability, then subtracted 100%. That gives a separate comparison from counting best prices.

BetRight had the lowest median full-field overround at 23.7%. Unibet had the highest at 32.7%. Each bookmaker’s figure comes from the same ten complete race markets.

Dot: median overround. Line: lowest to highest race overround. A lower figure means a smaller total implied-probability excess; it is not realized bookmaker profit.

A best price on one horse can sit inside a high-overround market. A comparison app needs both the runner price and the surrounding market.

All ten races. Every runner.

Six Mount Isa races and four Kyneton races. Open a race to see every matched win price and all five full-field overrounds. Blue prices are the highest in that row, including ties.

Put the comparison in your app

The useful part is the join: the same race, runner and market across several bookmakers. The racing odds API gives you race discovery, runner prices and streaming updates to build that comparison.

Start with a racecard and a current odds snapshot. Match runners by event ID and runner number, keep each bookmaker’s timestamps, and retain scratches separately from missing quotes. Our Python racing board has the working implementation.

For coverage and the racing endpoints, see the horse racing API. If you are calculating probabilities, the remove-vig guide explains why the quoted market percentage needs attention.

Method and data

We read the saved feed at 20 October 2026, 00:24 UTC. The sample contains ten Australian thoroughbred races, 71 active runners and 355 decimal win prices. Each runner has a price from the same five bookmakers.

Races were selected by complete-field bookmaker coverage, then bookmaker count and scheduled start. Price gaps, price leaders and results played no part. We matched runner numbers and normalized names, excluded scratched runners, and counted the verified Mount Isa/Mt Isa aliases once.

The five sources were selected for complete matching fields across all ten races. Betfair Exchange was excluded from this fixed-odds comparison. Ladbrokes and other available sources were outside this five-bookmaker sample.

Best-price frequency includes ties. Median price gaps weight each runner once. Median overround weights each race once. For the odds-range chart, each runner is grouped by its median price across all five bookmakers.

Source metadata timestamps were 10.7–24.3 minutes older than capture. They do not prove when a bookmaker last changed a price. This is one saved-feed observation across two meetings, rather than closing-price data or a measurement of movement before the start. We did not test whether these prices could be accepted for a stake or measure subsequent results.

The downloads contain the same 71 runners and 355 prices shown here, chart data and field definitions. Research by odds-api.net. Questions about the calculations? Contact us.