Loading...
Dutching replaces “which horse?” with “which group?” — and changes the maths entirely. Instead of selecting a single winner and staking on it, you back two or more horses in the same race, adjusting the stake on each so that you make the same profit regardless of which one wins. It is not a hedge. It is not a safety net. Done properly, dutching is a structured approach to races where your analysis identifies multiple viable contenders and you believe the combined probability of the group exceeds what the market implies.
The technique takes its name from old-school bookmaking practice (the “Dutch book”), and it has been used by professional punters for decades. Modern calculators and exchange betting have made it more accessible, but the underlying principle has not changed: you are betting on a set of outcomes rather than a single one, and the maths has to work for the set as a whole.
The core calculation in dutching is straightforward. You have a total stake budget for the race. You divide that budget across your selections in proportion to their odds, so that each selection would return the same profit if it wins. The formula for the stake on each horse is:
Stake = (Total Budget x Implied Probability of Horse) / Sum of Implied Probabilities of All Selections
In practice, you convert each horse’s decimal odds into an implied probability (1 / decimal odds), sum the probabilities of your selections, then allocate stakes proportionally. The key number is the sum of implied probabilities of your dutch group: if it is less than 100 percent (i.e. less than 1.0), you have a mathematically viable dutch. If it exceeds 100 percent, the overround on your group means you will lose money even if one of your selections wins.
Consider the market reality. According to data from Inform Racing, 75 to 80 percent of all winners come from the top five in the betting market. Dutching the top three or four — the runners your analysis identifies as genuine contenders — captures the majority of probable winners in any given race. The question is not whether the group includes the winner often enough, but whether the combined odds offer a positive expected return after stakes are allocated.
Free dutching calculators are available on most betting sites and apps. You input the odds of each selection and your total stake, and the calculator tells you how much to place on each horse and what the return will be. The arithmetic is handled for you — the skill lies in choosing which horses to include.
Dutching is not appropriate for every race. It works best in specific scenarios where the competitive profile of the race suits a multi-selection approach.
Open handicaps. Big-field handicaps — the 16-to-20-runner Saturday races at York, Newbury or Ascot — are the natural territory for dutching. These races are inherently difficult to solve with a single selection because the field is closely matched and the market struggles to separate the contenders. Your form study might identify four horses with strong claims, none of which you would confidently single out. Dutching the group lets you profit if any one of them wins, at a cost that is manageable if none does.
Races with a weak favourite. FlatStats data shows that favourites starting at 8/1 or longer win only about 8 percent of the time. In these races, the market is openly signalling uncertainty, and dutching two or three of the leading contenders at prices of 5/1 to 10/1 can produce a viable set of selections with a combined win probability that the market underestimates.
Where dutching does not work: small-field conditions races with a clear favourite. If one horse is 4/6 and the rest are 5/1 or bigger, there is no scope for a dutch that captures value. The favourite dominates the probability, and including it in a dutch at short odds drags the return down to unprofitable levels. Single-selection betting is the right approach when the market has a clear leader and you agree with the assessment.
The most common dutching mistake is including too many selections. Every horse you add to the dutch increases the combined implied probability of your group. Add enough horses and the group’s implied probability will exceed 100 percent, at which point you are guaranteed to lose money — the bookmaker’s margin has eaten your edge. Three or four selections is typically the maximum for a profitable dutch in a competitive handicap. Beyond that, you are paying more in overround than you can recoup in winning frequency.
The second mistake is dutching without an edge. Dutching is a staking method, not a selection method. If your four horses are chosen randomly rather than through genuine form analysis, dutching them does not create value — it just spreads your losses across a wider group. The edge must come from your assessment that the horses in your dutch have a higher combined chance of winning than the market implies. Without that assessment, dutching is just a more complicated way to lose at the same rate.
The third mistake is false confidence. Dutching feels safer than single-selection betting because you have more runners working for you. That feeling is deceptive. A dutch can still lose — in fact, in a 20-runner handicap with a four-horse dutch, you lose 16 times out of 20 in a random world. The emotional comfort of having multiple selections must not lead you to stake more than your bankroll management rules allow.
Imagine a 14-runner handicap where your form study has identified three strong contenders. Horse A is 5/1, Horse B is 7/1, and Horse C is 10/1. You want to stake a total of £20 on the race.
First, convert to decimal odds and implied probabilities. Horse A: 6.0 decimal, implied probability 16.7 percent. Horse B: 8.0 decimal, implied probability 12.5 percent. Horse C: 11.0 decimal, implied probability 9.1 percent. The combined implied probability of your group is 38.3 percent — well below 100 percent, so the dutch is viable.
Next, allocate stakes proportionally. Horse A gets £20 x (16.7 / 38.3) = £8.72. Horse B gets £20 x (12.5 / 38.3) = £6.53. Horse C gets £20 x (9.1 / 38.3) = £4.75. Total staked: £20.00.
If Horse A wins at 6.0, the return is £8.72 x 6.0 = £52.32, for a profit of £32.32. If Horse B wins at 8.0, the return is £6.53 x 8.0 = £52.24, profit £32.24. If Horse C wins at 11.0, the return is £4.75 x 11.0 = £52.25, profit £32.25. The profit is approximately equal regardless of which selection wins — that is the defining feature of a properly calculated dutch.
The key question is whether your three selections genuinely have a combined chance of winning that exceeds 38.3 percent. If your form study says yes — if you believe the trio covers the most likely winner in the race — the dutch is a sound bet. If you are not confident the group has a higher combined chance than the market implies, the right action is to pass the race entirely rather than dutch speculatively.