The essentials
Dutching means backing two or more mutually exclusive outcomes in the same market and dividing the total stake so that each covered winner produces the same gross return. Instead of staking the same amount on every selection, you stake more on shorter prices and less on bigger prices.
The arithmetic can make several separate bets behave like one combined position, but it cannot turn poor prices into good ones. A Dutch is attractive only when the combined prices of the outcomes you cover represent a better proposition than the probability of one of them winning. If the market margin is too large, or your assessment is wrong, spreading the stake merely distributes a negative expectation across several selections.
This guide explains the calculation, the meaning of the combined price, the difference between equal-return and equal-stake Dutching, and the practical risks that calculators often hide. Every monetary figure is a worked illustration, not a recommendation or a record of actual bets.
What Dutching actually does
Imagine a horse race in which you believe three runners account for most of the realistic winning chance. You could back just one and accept that either of the other two might beat you. Alternatively, you could back all three and apportion one total stake across them.
Only one horse can win the conventional win market. The selections are therefore mutually exclusive, which is essential. If two bets can both win, as may happen across related football or player markets, the standard Dutching calculation does not describe the complete position.
An equal-return Dutch aims to produce the same total return whichever covered selection wins. The net profit will also be equal if the same total stake is at risk in every outcome. If an uncovered selection wins, every leg of the Dutch loses.
| Approach | How stakes are divided | Result if a covered selection wins | Main weakness |
|---|---|---|---|
| Equal stake | Same cash stake on each selection | Return varies with the winning odds | Bigger prices dominate the result |
| Equal return | Stakes vary according to price | Gross return is approximately level | Rounding can create small differences |
| Target profit | Stakes are set around a desired net result | Profit can be level if fully funded | The required outlay may be larger than expected |
The important phrase is covered selection. Dutching increases the number of outcomes that can return money, but every outcome outside the group remains a losing result.
The core Dutching formula
The cleanest calculation uses decimal odds. First convert each price into its reciprocal:
Reciprocal = 1 divided by decimal odds
Add the reciprocals of all covered selections. This total is the combined implied probability represented by the available prices. The effective decimal price of the Dutch is the reciprocal of that total:
Combined Dutch price = 1 divided by the sum of all reciprocals
To divide a total stake, multiply the total stake by each selection's reciprocal, then divide by the sum of all reciprocals:
Selection stake = total stake multiplied by selection reciprocal divided by reciprocal sum
The same calculation can be expressed more directly:
Selection stake = total stake divided by selection odds divided by reciprocal sum
If the reciprocal sum is below 1.00, the available prices imply a combined chance below 100%, so the Dutch produces a positive return if one of the covered outcomes wins. If the sum reaches or exceeds 1.00, an equal-return Dutch cannot produce a positive profit before any commission or other costs.
Complete three-selection example
Suppose three fictional runners are available at decimal odds of 3.50, 5.00 and 7.00. You decide to examine a total outlay of £30.
| Selection | Decimal odds | Reciprocal | Raw share of total |
|---|---|---|---|
| Runner A | 3.50 | 0.285714 | 53.19% |
| Runner B | 5.00 | 0.200000 | 37.23% |
| Runner C | 7.00 | 0.142857 | 26.59% |
| Reciprocal sum | - | 0.628571 | - |
The percentages in the final column must be normalised because the raw reciprocals add to 62.857%, not 100%. Divide each reciprocal by 0.628571. The resulting stake shares are approximately 45.45%, 31.82% and 22.73%.
| Selection | Stake before rounding | Practical stake | Gross return if it wins | Net result after £30 total stake |
|---|---|---|---|---|
| Runner A at 3.50 | £13.636 | £13.64 | £47.74 | £17.74 |
| Runner B at 5.00 | £9.545 | £9.55 | £47.75 | £17.75 |
| Runner C at 7.00 | £6.818 | £6.81 | £47.67 | £17.67 |
| Total | £30.000 | £30.00 | Approximately £47.72 | Approximately £17.72 |
The small variation comes from rounding stakes to pennies. The theoretical combined price is:
1 divided by 0.628571 = 1.5909
Multiplying £30 by 1.5909 gives a theoretical return of £47.73 and a theoretical profit of £17.73. The Dutch behaves like a single bet at about 1.59 on the proposition that A, B or C wins.
This is not free money. Any other runner wins the race often enough to justify the apparent 1.59 price if the market is efficient. The calculation equalises the pay-off; it says nothing about whether the group has a genuine 62.86% chance.
Combined price versus true probability
This is the decision that matters. The available prices imply that the covered group has a 62.86% chance before any adjustment for market margin. Your own research might estimate the group at 66%, 60% or something else entirely.
If you estimate 66%, fair decimal odds for the group are 1 divided by 0.66, or 1.515. A combined market price of 1.591 would be bigger than your assessed fair price, creating positive theoretical expected value under your estimate.
If you estimate only 60%, fair odds are 1.667. Taking 1.591 would then be too short, even though three selections are covered and the bet may win relatively often.
| Your estimated chance of any covered winner | Your fair combined odds | Available combined odds | Theoretical view |
|---|---|---|---|
| 58% | 1.724 | 1.591 | Price is too short |
| 62.86% | 1.591 | 1.591 | Break-even before costs |
| 66% | 1.515 | 1.591 | Positive expectation under the estimate |
| 70% | 1.429 | 1.591 | Larger apparent edge, if the estimate is sound |
The phrase under the estimate is crucial. A precise spreadsheet cannot rescue an inaccurate probability. Read our guides to implied probability and expected value in betting before treating a combined price as evidence of value.
How bookmaker margin affects a Dutch
Every selection price should be considered within the whole market. Bookmaker odds commonly imply more than 100% when all possible outcomes are converted into probabilities. That excess is the overround, although the margin may not be distributed evenly.
Dutching several runners does not remove the bookmaker's margin. It can concentrate it if your chosen selections are priced less generously than the rest of the field. Comparing the combined implied probability with a fair, margin-free view of the same group is more useful than celebrating a high strike rate.
Price shopping matters because small changes compound across the Dutch. In the example above, moving Runner A from 3.50 to 3.30 increases its reciprocal from 28.57% to 30.30%. The reciprocal sum rises, the combined price falls and the potential return on the same £30 declines.
Use the actual accepted odds, not an advertised figure that was unavailable for your intended stake. Our guide to how betting odds work explains why total return, profit and implied probability must be kept separate.
When Dutching can make analytical sense
Dutching is a staking structure rather than a selection method. It can be useful when your analysis concerns a group of outcomes and you can price that group more confidently than any one member.
In horse racing, you might conclude that a particular pace scenario strongly favours three prominent runners but be unsure which one will finish best. In a golf tournament, your model might identify several players whose combined winning probability exceeds the probability implied by their available prices. In a correct-score football market, you might cover a tightly defined group of scorelines, although settlement details and correlation assumptions still require care.
The method is not justified merely because several selections look appealing. A sensible process asks:
- Are the outcomes in the same market and mutually exclusive?
- What is the reciprocal sum at prices genuinely available now?
- What probability do I assign to the covered group?
- How was that estimate produced and how uncertain is it?
- What happens if every uncovered outcome is treated as one losing side of the bet?
- Do rounding, commission, deductions or settlement rules change the result?
Covering more runners can make a bet feel safer because wins may occur more frequently. Yet the combined price shortens as each new selection is added. The price, not the number of names on the slip, determines the break-even requirement.
Equal profit is not equal value
A Dutch calculator is designed to balance returns. It does not identify mispriced selections. Two poor-value bets can be combined into one poor-value position with beautifully equal profits.
Suppose your analysis values two mutually exclusive outcomes at 30% and 20%, giving the group a 50% true chance. Fair combined odds would be 2.00. If the available prices produce a reciprocal sum of 0.55, the combined Dutch price is only 1.818. Equalised stakes cannot change the fact that you would be accepting 1.818 about a proposition you believe should be 2.00.
This distinction also explains why a high strike rate is not proof of success. A Dutch priced at 1.40 needs to win more than 71.43% of the time before costs merely to break even. Winning seven bets in ten would still lose money at that price.
Common Dutching mistakes
Using equal stakes by accident
Backing three runners for £10 each is not an equal-return Dutch unless their odds are identical. A winner at 3.00 returns £30, while a winner at 8.00 returns £80. Decide whether the objective is equal stake or equal return before calculating anything.
Mixing markets that can both win
Standard Dutching assumes one covered outcome settles as the winner and the others lose. Combining a match winner with a player prop, or a horse to win with the same horse to place, creates a different payoff structure. Model every possible joint outcome rather than applying a simple reciprocal formula.
Ignoring dead heats and deductions
Horse racing dead-heat rules can reduce the winning portion of a bet. Rule 4 deductions may reduce returns after a non-runner. Golf ties can affect place markets. Read the operator's settlement terms before treating the displayed price as the final pay-off.
Treating an uncovered outcome as unlikely rather than costly
Every excluded runner or scoreline belongs to the losing side of the position. If the Dutch returns £47.73 from £30 when covered, it loses the full £30 when uncovered. Analyse both amounts and their probabilities.
Recalculating after a price moves without checking the total
If one selection shortens or drifts before bets are placed, all stakes should be recalculated. Mixing old stake amounts with new prices breaks the equal-return structure.
Chasing losses with a larger total stake
Increasing the next Dutch to recover a previous loss does not improve the probabilities or prices. It simply puts more money at risk. A staking progression should never be confused with evidence of value.
A practical calculation checklist
| Step | Calculation or check | Record to keep |
|---|---|---|
| 1 | Confirm one market and mutually exclusive outcomes | Event, market and settlement rules |
| 2 | Capture decimal prices | Operator, time and accepted odds |
| 3 | Calculate each reciprocal | 1 divided by each price |
| 4 | Add the reciprocals | Combined implied probability |
| 5 | Calculate combined odds | 1 divided by reciprocal sum |
| 6 | Estimate the group's true probability | Method, evidence and uncertainty |
| 7 | Divide the total stake | Normalised reciprocal shares |
| 8 | Check every payoff | Covered and uncovered results |
| 9 | Apply real-world costs | Commission, deductions and rounding |
| 10 | Review after settlement | Return, profit or loss and process notes |
The BetOwl bet calculator can help check returns, but keep the underlying formula in your notes. Understanding the mechanics makes it easier to spot a wrong input or a calculator that assumes different settlement rules.
Dutching on a betting exchange
Exchange markets can offer more granular decimal prices, but available liquidity matters. A headline price may be available for only part of the required stake. If one leg is partially matched and the others are fully matched, the final position will not have equal returns.
Commission can also reduce the winning market result. The exact charging method depends on the exchange and customer account, so a before-commission calculation should not be presented as the final net profit. Record each matched portion, its price and any unmatched balance.
Do not confuse backing several selections with laying the remainder of the field. The positions can sometimes produce similar gross exposure, but exchange liability, commission and settlement mechanics differ. Calculate the complete payoff table rather than assuming equivalence.
Bankroll and responsible use
Dutching can create an illusion of protection because several outcomes are covered. The entire total stake is still exposed whenever an uncovered result occurs. A shorter combined price may also encourage larger stakes because the group appears more likely to win.
Set the maximum total outlay before dividing it among selections. Do not calculate the stake required to hit a desired income target and then assume that amount is affordable. The sensible direction is budget first, allocation second.
Betting should remain optional entertainment for adults aged 18 and over. Never use money needed for bills, borrowing or essential spending. Use account limits, time-outs or self-exclusion if betting is becoming difficult to control. The responsible gambling guide explains practical safeguards and UK support options.
The final verdict
Dutching is useful mathematics, not a shortcut to profit. It converts several mutually exclusive bets into one combined proposition and can equalise the return across the covered outcomes. That makes the exposure clearer and can suit analysis built around a group rather than a single selection.
Its limitation is equally clear: stake allocation cannot create value. The quality of the position still depends on the true combined probability, the prices actually obtained, the market margin, costs and settlement rules. Calculate the Dutch only after you have priced the group. If the combined odds are too short, a neat spread of stakes remains a bad bet.
Frequently asked questions
What is Dutching in betting?
Dutching is backing two or more mutually exclusive outcomes in one market and dividing a total stake between them. An equal-return Dutch uses different stakes so that each covered winner produces approximately the same gross return.
Can Dutching guarantee a profit?
No. Every outcome outside the covered group can still lose the entire total stake, and the covered prices may represent negative expected value. Dutching changes the payoff distribution, not the underlying chance or quality of the prices.
How do I calculate Dutching stakes?
Convert each decimal price to a reciprocal, add the reciprocals, then divide each reciprocal by the total to obtain its stake share. Multiply each share by the total amount you have decided is affordable to risk.
Is Dutching the same as an accumulator?
No. A conventional accumulator requires every leg to win. A Dutch usually covers several mutually exclusive outcomes in one market and needs only one covered selection to win.
Does adding more selections make a Dutch safer?
It increases the range of covered outcomes but also shortens the combined price and can add poor-value selections. Risk should be judged from the complete payoff and true probability, not simply from how many outcomes appear on the slip.