Dutching for Matched Betting: When to Skip the Lay

Dutching is named after Arthur 'Dutch' Schultz, a US Prohibition-era gambler whose trick was backing several horses in a single race in proportions that produced the same net return whoever won. The same mathematics applies to any market with mutually exclusive outcomes: tennis matches (two outcomes), football match results (three outcomes), horse racing markets (any number of outcomes).

Mechanically: you back outcome A for stake S_A, outcome B for S_B, and so on, choosing the stakes such that the gross winnings from each winning leg are equal. If you do this across a market where the implied probabilities sum to less than 1 (i.e. the bookmaker's overround is below 100%), you have a guaranteed profit - that is the rare case of pure arbitrage. More commonly the implied probabilities sum to slightly more than 1, in which case you have a guaranteed small loss instead - which is the qualifying-loss equivalent that matched bettors recognise.

For matched betting purposes, dutching is most often used to convert a qualifying bet (or a free bet) without ever touching the exchange. Instead of backing one horse at Sky Bet and laying it at Betfair, you back multiple horses across one or several bookmakers in stakes that guarantee a small loss regardless of result.

The core calculation is straightforward. For each outcome, compute the implied probability as the inverse of the decimal odds. Sum those probabilities to get the market total. If the total is less than 1, you have arb; if it is greater than 1, you have a guaranteed qualifying loss equal to the overround.

To stake equally across outcomes - so that any winning leg returns the same gross payout - your stake on outcome i is proportional to (1 / odds_i). Concretely, if you want a total bankroll B distributed across n outcomes, the stake on outcome i is:

stake_i = B × (1/odds_i) / Σ(1/odds_j)

The guaranteed payout from any winning leg is then B / Σ(1/odds_j). Subtract B to get the profit (or loss).

In practice you do not crank this through by hand on every dutching bet. The community calculators do the maths instantly - see our lay stake calculator guide for the calculators that cover both laying and dutching, and the dedicated dutching tools listed at the end of this piece.

Default to exchange laying when the market is liquid (Premier League match odds, major horse racing meetings, big tennis matches). The exchange overround is typically smaller than the dutching overround across multiple bookmakers, and the back-and-lay structure is simpler to reason about.

Default to dutching when one or more of the following apply:

• Thin exchange liquidity. Lower-league football, niche tennis markets, low-grade horse meetings, and most non-mainstream sports have an exchange lay price so wide that the qualifying loss kills any expected value. Dutching across bookmakers - where each bookmaker prices the market based on their model rather than a market clearing price - often beats the exchange.
• Money-back specials. When the refund triggers on a specific outcome (e.g. 'money back if your first goalscorer scores last too', 'money back if 0-0'), dutching is often the only way to capture both the qualifying outcome and the refund-eligible outcome cleanly.
• Free bet conversion in low-liquidity markets. A stake-not-returned free bet that has to be used on a specific market (horse racing only, football only, etc.) can sometimes be converted more efficiently by dutching across bookmakers than by backing high and laying high at the exchange.
• Avoiding exchange-account profile. Some matched bettors prefer to keep their exchange volume low to avoid being flagged as a heavy-promotion account by their banking relationships or for reasons of personal preference. Dutching keeps the entire flow on bookmaker accounts.

For a refresher on the standard back-and-lay free-bet conversion methods that dutching complements rather than replaces, see our free bet conversion methods guide.

Money-back specials are bookmaker offers where the qualifying bet is refunded if a specific event occurs - typical examples include 'money back if the match is a draw' on football, 'money back if your horse finishes second' on racing, or 'money back as a free bet if your first scorer scores last too'. The standard back-and-lay approach struggles here because the refund pays out on a specific outcome the exchange lay-off doesn't capture cleanly.

Dutching is often the elegant solution. The play: back the qualifying market at one bookmaker (say, back a horse to win), then back the refund-trigger outcome at the same or different bookmaker (back the second-favourite to win, which would trigger the refund if your first horse finished second to it). Size the stakes so that whichever outcome wins, your net position is acceptable, and the refund triggers a free bet worth more than your worst-case dutching loss.

The expected value here can be substantial. A money-back special that turns into a 70%-conversion-rate free bet is effectively a 'free hit' if your dutching loss is less than 30% of the refund value. The volume of these offers is highest during football season and major racing festivals.

The strongest case for dutching over laying is when the exchange just doesn't have enough money in the market to lay at a sensible price. This applies most clearly to:

• Lower-league football. League Two, National League, Scottish lower divisions, mid-week European competitions. Match-odds liquidity on Betfair often drops to triple-digit pounds in the hours before kick-off, and the lay price can sit 15–20% above the back price.
• Niche tennis. Challenger and Futures tour matches frequently have lay prices 10% above back prices, sometimes more.
• Most racing outside the UK and Ireland. French, American, Australian, and Hong Kong racing is supported by every major UK bookmaker, but exchange liquidity outside the mainstream UK meetings is often non-existent.
• Specials and proposition markets. 'Both teams to score and over 2.5 goals', 'first goalscorer', 'method of dismissal in cricket' - these specials are priced by every bookmaker but rarely have any exchange depth.

In all of these cases, comparing the dutching overround across two or three bookmakers (e.g. Sky Bet vs William Hill vs Bet365) often produces a tighter combined market than the single exchange lay price. Comparison sites like Oddschecker make this easy: scan for the best price on each outcome, then run the dutching calculation across the cheapest combination.

Manual dutching across multiple bookmakers is slow - by the time you have computed stakes for three outcomes and placed the bets, prices may have moved. Three categories of tooling exist to help:

• Free standalone calculators. Most matched-betting sites publish a dutching calculator - you punch in odds and total stake, the calculator outputs per-outcome stakes. AceOdds, MatchedBets, and OddsMonkey all have free versions. Good for occasional dutching; tedious for daily volume.
• Paid matched-betting platforms. OddsMonkey and Outplayed include dutching matchers that scan multiple bookmakers' prices simultaneously and surface low-overround dutching opportunities. RebelBetting is the historical premium option for proper sportsbook arbing - closest to a 'just give me the arb' tool but at a higher monthly cost. The marginal value over the free calculators is real if you are doing daily volume.
• Spreadsheet templates. A simple Google Sheet with implied-probability formulas and stake outputs handles the maths for the cost of nothing. Most experienced matched bettors keep one for ad-hoc dutching outside their normal matcher tool. Useful when the matcher misses a market that you have spotted manually.

Whichever tool you use, sanity-check the stakes before placing the bets: a small typo in the odds entry can produce stakes that are catastrophically wrong.

Five mistakes account for almost all dutching losses beyond the expected overround cost:

1. Not accounting for the dead-heat rule on each-way bets. Each-way matched-betting overlaps with dutching when you back the win at one bookmaker and the place at another. See our each-way matched betting guide for the dead-heat rules - they materially change the dutching arithmetic.
2. Stale odds. The matcher tool shows the odds it had when it last scraped the bookmaker. By the time you click through and place the bet, the price may have moved. Always confirm the live odds at each bookmaker before placing each leg.
3. Placing legs in the wrong order. If one of your legs is at a bookmaker with weak limits, place that one first. If it gets restricted before placement, you can adjust stakes at the other bookmakers; you cannot adjust the first leg once placed.
4. Stake size errors from rounding. Decimal odds with three or four significant figures need correspondingly precise stake calculations. Rounding £41.06 to £41 on one leg and £30.79 to £31 on another can produce a £1+ swing in net result.
5. Forgetting the bookmaker's payout cap. Maximum-payout limits on bookmaker accounts apply per bet, not across your dutching set. A big-stake play across multiple bookmakers should check that no single leg payout exceeds the bookmaker's cap.

Dutching does not replace exchange laying as the core matched-betting technique - for liquid markets, the exchange will almost always be more efficient. But for the offers that lay-off cannot capture cleanly, for the markets where liquidity is light, and for the bookmaker promotions that pay on specific outcomes, dutching is a strictly better tool. Adding it to your toolkit broadens the set of promotions you can extract value from, and slightly reduces the per-bookmaker pace at which restrictions arrive.