Best Bookmakers for Finding Arbitrage Bets in Australia

The bookmakers that appeared most

We counted a bookmaker once per distinct opportunity, even if its name appeared on more than one selection. Percentages overlap because an arbitrage opportunity normally involves two bookmakers.

Sportsbet's lead was broad, not universal. PointsBet appeared more often in handicaps and team totals. Unibet led moneyline, total-corners and team-total-corners appearances. PlayUp and TAB tied for the most three-way moneyline appearances among the books in that market.

Frequency tells you where price disagreements occurred. It does not tell you which bookmaker is easiest to use. Deposit methods, identity checks, limits and account treatment need a separate comparison.

Five bookmakers covered one side. Not both.

The top five—Sportsbet, PointsBet, Unibet, Ladbrokes and Dabble—appeared on at least one side of 292,800 opportunities. That is 70.2% of the archive.

Both sides sat inside that same five-bookmaker group only 116,205 times, or 27.9%. The practical takeaway is simple: a tight shortlist helps, but a wider roster finds combinations the shortlist cannot complete.

Of the 417,088 records, 414,257—99.3%—contained two distinct bookmaker names. We kept the remaining 2,831 in the archive-level rankings, but they should not be treated as placement-ready two-bookmaker opportunities.

Totals and handicaps generated 55.8%

Totals produced 133,816 opportunities. Handicaps produced 98,767. Combined, those two market types accounted for 232,583 records—more than the next eight market types combined.

The top 10 market types represented 82.1% of the archive. That concentration matters. If your arbitrage finder misses totals or handicaps, it is not missing a niche. It is missing most of the market.

The result also kills a common assumption: moneyline was not the main source of arbs. Exact lines created far more opportunities. Over 214.5 and under 214.5 must be matched exactly; comparing a different line creates a middle, a gap or a straight mismatch—not a clean arb.

The typical percentage was 1.23%

Older archive rows did not retain a usable percentage field, so this part uses 133,770 valid latest-price snapshots from the newer archive only. We removed missing, non-finite, stale and sentinel values, then kept positive projected margins no greater than 15%.

10th percentile

0.18%

Median

1.23%

90th percentile

6.62%

95th percentile

9.77%

The middle 80% ran from 0.18% to 6.62%. The full valid range was 0.001% to 15%, but the endpoints are a poor description of a normal opportunity. The median is the number worth remembering.

Three-way moneyline produced the largest median in this filtered snapshot, but it had a much smaller sample and a distribution close to our 15% quality ceiling. Treat that result as a feature of the observed data, not a promise that football moneylines will return 9.76%.

These percentages are projected margins from the listed prices. They are not customer results. Odds can move, a stake can be rejected, settlement rules can differ and one leg can be voided.

NBA and MLB led the leagues

NBA generated 81,312 records, followed by MLB on 66,653. Australia's NRL and AFL ranked third and fourth. Together, the top 10 named leagues accounted for 67.6% of every opportunity.

League volume is seasonal. NBA and MLB were active through large parts of this window, while other competitions had shorter or quieter periods. A six-month count is useful for choosing coverage; it is not a permanent league ranking.

We excluded 35,625 records with no league label from the league ranking. They remain in the overall, bookmaker and market totals.

A complete $1,000 arbitrage example

Assume Sportsbet lists Team A at 2.10 and PointsBet lists Team B at 2.10 in a hypothetical two-way market. The selections oppose each other and use the same settlement rules.

Now move one price. If the second price drops from 2.10 to 1.95 after the first $500 bet is accepted, another $500 no longer balances the result. Team A still returns $1,050. Team B returns $975, leaving a $25 loss on that outcome.

That is the execution risk hidden by an archive row. Recheck both prices and recalculate the split before confirming the first stake.

How WagerWise finds arbitrage bets

The archive came from WagerWise's own arbitrage records. The live WagerWise arbitrage finder does the same basic job in real time: it compares equivalent markets across bookmakers and surfaces combinations where the best listed prices imply less than 100%.

Evidence behind it

417,088 archived opportunities

What it compares

Bookmakers, markets and prices

What it calculates

Projected return and stake split

An arbitrage exists when the best available odds across every outcome imply less than 100%. You then split the total stake so the payout is balanced across those outcomes.

1. Compare more than one bookmaker. Our data shows why: no single pair dominated the archive.
2. Match the exact market and line. Team, period, player, total and settlement rules must agree.
3. Check every price again. A stale quote can turn a projected profit into a loss.
4. Calculate the stake split. Use the WagerWise arbitrage calculator before placing either side.
5. Confirm both bets. Do not assume the second stake will be accepted because the first one was.

Manual comparison works for a small number of events. WagerWise handles the wider scan and puts the bookmaker, market, prices, projected return and stake split together. It does not remove execution risk: prices can move and a bookmaker can reject the second stake.

The arbitrage betting guide explains the maths and the practical risks in more detail.

How we built the study

We pulled every distinct archived arbitrage record first seen from 14 March to 14 September 2026, using Brisbane dates. The newest retained version of each bet ID won during deduplication.

Known limitations

• The archive has a known nine-day source handover gap from 11 to 19 April 2026.
• 14 September was still in progress when the export was generated.
• Bookmaker appearances measure opportunity coverage, not accepted stakes, limits, withdrawals or account tolerance.
• Bookmaker names reflect the archive during the study window. Current availability and legal access can change.
• The archive records detected opportunities. It does not prove a customer placed both bets or earned the projected margin.
• League labels were missing on 35,625 records. Those rows were excluded only from the league table.
• Percentage analysis uses the newer archive because the legacy rows did not contain a comparable percentage field.
• The two archive systems did not retain a comparable opportunity-lifetime series, so we did not publish detection-speed or lifespan claims.

Source: WagerWise arbitrage opportunity archive. Export generated 14 September 2026. Counts are exact; displayed percentages are rounded.

Questions the data can answer