Dutching Explained: Backing More Than One Horse in the Same Race
Sometimes the honest answer to "which one wins?" is "one of these three". Dutching is the technique for betting exactly that opinion: splitting your stake across several horses in proportions that return the same profit whichever of them wins.
TL;DR: Dutching means backing two or more horses in one race with stakes weighted by their odds, so any of your selections winning returns identical profit. It's not a system and it doesn't manufacture value — your dutched book wins if the combined true chance of your horses exceeds their combined implied probability. Where it shines: races where you can confidently rule most of the field out but can't split the few you've ruled in.
The basic idea
Backing one horse is a claim: this one wins. Dutching is a broader claim: the winner comes from this group, with stakes arranged so you're indifferent about which member of the group obliges.
The staking rule: each horse's stake is proportional to its implied probability (1 ÷ decimal odds). Shorter-priced selections take more of your total stake, longer prices take less, and everything equalises.
A worked example with a £30 total stake across three horses:
| Horse | Decimal odds | Implied prob. | Stake | Return if it wins |
|---|---|---|---|---|
| A | 4.0 | 25.0% | £15.79 | £63.16 |
| B | 8.0 | 12.5% | £7.89 | £63.12 |
| C | 10.0 | 10.0% | £6.32 | £63.20 |
The arithmetic for the general case: divide each horse's implied probability by the sum of your selections' implied probabilities, and that fraction of your total stake goes on that horse. Our dutching calculator does it in a second; a spreadsheet does it forever.
What dutching is really doing
It helps to see the equivalent single bet. In the example above, you've built a bet with a 47.5% implied chance paying out at effective odds of 2.11 (£63.16 ÷ £30). Dutching combined three prices into one synthetic price.
Which exposes the truth about the technique: dutching is a stake-allocation tool, not an edge. The dutched bet is value if and only if your group's true combined win chance beats 47.5%. Rate the trio's real chance at 55% and this is an excellent bet. Rate it honestly at 45% and you've tidily organised a losing wager. The expected value arithmetic is untouched by how cleverly the stakes were split.
The genuine advantages are practical:
- It matches how race analysis actually ends. Serious form study often concludes "it's between these three" with real confidence, while the choice among them is a coin toss. Dutching bets the strong conclusion instead of forcing a weak one.
- It removes the worst outcome in betting psychology: doing the analysis, picking the wrong one of your two fancies, and watching the other one win. Over a season that outcome damages decision-making more than most people admit.
- It handles the overround honestly. You're consuming several slices of the bookmaker's margin at once, which is why the prices you take matter double when dutching — see below.
The structural disadvantage: your strike rate rises but your payout compresses. High combined probabilities produce short synthetic odds; dutch 80% of the market and you're effectively backing a 1/4 shot. Dutching a group whose members are individually poor value compounds bad prices into one guaranteed-mediocre bet.
Where dutching earns its keep
Chaotic big-field handicaps. Twenty runners, five of which you genuinely can't separate, the rest of which you're happy to oppose. Dutching the five at big prices can build a synthetic bet at fair odds with a plausible edge. This is dutching's natural habitat, and it's the same territory as big-field each-way value, approached from a different angle.
Strong negative views. Sometimes the sharpest opinion in a race is "the favourite is wrong". Dutching the credible alternatives is the back-side expression of that view (the direct expression is laying the favourite — compare the effective odds both ways; the better instrument varies race by race with the exchange spread and commission).
Prices scattered across bookmakers. Dutching three horses doesn't mean betting with one firm. Taking each selection's best available price (different books for different horses) can add several points of combined margin versus dutching at a single bookmaker's prices. This is mechanical, free improvement, and exactly what an odds comparison screen is for. The difference between dutching three horses at one book and at three books' best prices is routinely the difference between a losing and winning synthetic price.
When not to dutch: when one of your group is clearly value and the others are just company. Adding fairly-priced or under-priced horses to a value selection dilutes the edge. Dutching spreads stake, and edge should be concentrated where it's largest — the same staking logic that governs any bet.
Dutching vs each-way, and other comparisons
| Approach | The claim it bets | Pays when |
|---|---|---|
| Single win bet | "This horse wins" | Your horse wins |
| Each-way | "This horse runs well" | Wins or places |
| Dutching | "The winner is in this group" | Any selection wins |
| Lay the favourite | "That horse doesn't win" | Anyone else wins |
A second worked example: dutching against a favourite
The negative-opinion use case, with numbers. An odds-on favourite at 1.8 looks vulnerable to you, and three of the other five runners are the credible beneficiaries, priced 5.0, 7.0 and 11.0.
Their implied probabilities: 20% + 14.3% + 9.1% = 43.4% combined. Dutching £30 across them in proportion (about £13.83, £9.88 and £6.29) returns roughly £69 whichever wins: a synthetic price of 2.30 on "one of these three beats the favourite (and the other two)".
Now the check that decides whether to bet. The favourite's 1.8 implies 55.6%. If your reading says its real chance is 45%, then roughly 55% of the probability belongs to the rest, and if your three collect most of it (say 48%), your 43.4% synthetic bet is value. If instead the two outsiders you excluded are live, your group's true share might be only 40%, and the identical dutch is a losing bet. Same technique, opposite outcomes, decided entirely by the probabilities you feed it.
Worth comparing before striking it: laying the favourite at 1.8 risks £24 liability per £30 stake-equivalent and wins if anyone beats it, including the outsiders you excluded. The lay is the cleaner instrument when your view is purely anti-favourite; the dutch pays better when you can genuinely narrow the alternatives.
Frequently asked questions
How do I calculate dutching stakes by hand?
Convert each price to implied probability (1 ÷ decimal odds). Divide each horse's probability by the group's total probability. Multiply each fraction by your total stake. Equal profit follows automatically.
Can I dutch to a target profit instead of a set total stake?
Yes — decide the profit you want, and the same proportions determine what total stake achieves it. Calculators offer both modes.
Does dutching work on exchanges?
Fully, with commission trimmed off returns. Mixed dutching — some legs at bookmakers, some at exchange prices — is common, since you should be taking the best price per horse wherever it lives.
How many horses is too many?
There's no rule, but the more you add, the shorter your synthetic price and the more overround you swallow. Past three or four selections, check the effective combined odds and ask whether you'd back a single horse at that price with the group's combined chance.
Is dutching related to arbitrage?
Mechanically similar (stake-splitting for equalised outcomes), but arbitrage covers every runner at prices summing under 100% — a guaranteed profit that bookmaker margins almost never permit. Dutching covers part of the field and stays a genuine bet.
What happens to my dutch if one horse is a non-runner?
That leg is voided (or Rule 4'd if others withdraw) and the equal-profit structure breaks — your remaining stakes no longer balance. Recalculate and top up if the race still justifies it.