Dutching is what happens when you fancy more than one horse in a race and want to guarantee the same profit regardless of which one wins. Instead of picking a single selection and hoping for the best, you split your budget across two, three, or more runners, adjusting the stake on each so that every winning outcome returns the same amount. It is not hedging. It is not covering your bases out of indecision. It is a structured approach to a race where the field is so large and the outcome so uncertain that backing a single horse feels like leaving money on the table.
The Grand National, with 34 runners and an average winning price of roughly 18.7/1 over the past decade, is one of the few races in the calendar where dutching a small group of selections can be both mathematically sound and practically sensible.
What Dutching Actually Is
The term comes from the Dutch book, a concept in probability theory where a set of bets guarantees a profit regardless of the outcome. In racing, dutching refers specifically to backing multiple horses in the same race with proportionally adjusted stakes so that each winning outcome delivers an equal return.
The logic is easiest to see with two horses. If you like Horse A at 10/1 and Horse B at 14/1, a straight ten-pound bet on each would return different amounts depending on which wins. Dutching adjusts the stakes so the returns match. You bet more on the shorter-priced horse and less on the longer-priced one, allocating your total budget in inverse proportion to the odds.
The key principle is that your total stake across all selections must be less than your projected return from any single winner. If it is not, the dutch is unprofitable. This constraint limits how many horses you can realistically dutch in a single race: the more runners you include, the smaller each individual stake becomes, and at some point the combined stakes exceed the potential return from the shortest-priced selection.
The Calculation Method
The formula for dutching is based on implied probabilities. Each horse’s odds can be converted into an implied probability by dividing 1 by the decimal odds. A horse at 10/1 in fractional odds is 11.0 in decimal. Its implied probability is 1 divided by 11, which equals 0.0909, or about 9.09 per cent.
To dutch a group of horses, add up their implied probabilities. If the total is less than 1 (100%), the dutch is profitable — if one of your selections wins, your return exceeds your total stake. If the total exceeds 1, the dutch guarantees a loss no matter which horse wins, and you should not proceed.
The stake for each horse is calculated as its implied probability divided by the sum of all implied probabilities, multiplied by your total budget. For a twenty-pound budget split between Horse A at 10/1 (decimal 11.0) and Horse B at 14/1 (decimal 15.0): the implied probability of A is 0.0909, of B is 0.0667, and their sum is 0.1576. The stake on A becomes 11.54, the stake on B becomes 8.46, and the total is 20 pounds exactly.
If A wins at 10/1, the return is 11.54 multiplied by 11.0, which gives 126.94. If B wins at 14/1, the return is 8.46 multiplied by 15.0, giving 126.90. The returns are effectively equal. Your profit from a twenty-pound outlay is approximately 107 pounds regardless of which horse wins.
The average Grand National winner over the past decade has started at approximately 18.7/1. At those odds, a two-horse dutch involving selections in the 12/1 to 25/1 range typically produces a combined implied probability well under 20 per cent, leaving substantial room for profit.
A Grand National Worked Example
Suppose you have a thirty-pound Grand National budget and three horses you believe have strong claims: one at 12/1, one at 16/1, and one at 20/1.
Convert to decimal odds: 13.0, 17.0, 21.0. Implied probabilities: 0.0769, 0.0588, 0.0476. Sum: 0.1833. Stakes: Horse at 12/1 gets 12.59, horse at 16/1 gets 9.62, horse at 20/1 gets 7.79. Total: 30 pounds.
Return if any wins: approximately 163.60. Profit: approximately 133.60 on a thirty-pound outlay. The key number is the combined implied probability of 18.33 per cent — well under 100 per cent, confirming the dutch is viable.
Since 1999, only six pre-race favourites have won the Grand National. The rest went to horses at double-digit prices, often from the 14/1 to 33/1 range. That pattern, documented across decades of results tracked on sites like GrandNational.fans, is precisely what makes dutching attractive. If favourites rarely oblige, spreading your money across several live contenders at bigger prices increases the probability that one of your selections is the horse that does.
Limits and Caveats
Dutching is not a free lunch. The primary limitation is that it only pays when one of your selected horses wins. If none of them win — which, in a 34-runner steeplechase, is the most likely single outcome for any small group of selections — you lose your entire combined stake. Dutching increases your probability of having a winner compared to backing one horse, but the probability of losing is still the dominant outcome.
The second limitation is the each-way question. Pure dutching is a win-only strategy. If you want place protection, you need to either dutch each-way (which doubles your total outlay and complicates the maths significantly) or accept that the dutch is a win-or-nothing proposition. For the Grand National, where each-way is the default bet type and the place returns are often substantial, this is a genuine trade-off. Dutching three horses to win at a combined thirty-pound stake offers no return if one finishes second. Three separate five-pound each-way bets on the same horses, costing the same thirty pounds, would return a place payout on any that finished in the top four or five.
The third caveat is practical: bookmakers notice dutching. If you consistently dutch across multiple accounts or place suspiciously proportional stakes, some operators may restrict your account. For a once-a-year Grand National punt this is not a realistic concern, but for regular bettors who dutch as a habitual strategy, account restrictions are an occupational hazard.
Dutching suits the Grand National bettor who has done enough homework to narrow the field from 34 to three or four genuine fancies but cannot separate them further. It converts uncertainty into structure, spreading your budget so that any of your shortlisted horses winning produces the same satisfying return. The maths is mechanical. The hard part, as always, is picking the right horses in the first place.