Picture a moment most racing fans know. You have been through the card for twenty minutes. Three horses look live. One is the obvious favourite, one is a fancied stayer from an in-form yard, and one is a price that has caught your eye. You pick the favourite because you are meant to find one winner. The 14/1 shot wins. The 7/1 stayer is second.
Tipster services often frame the job as finding the winner. Backing more than one horse in a race can feel like doubt, or greed. It is neither. It is a legitimate way to bet when your own view honestly says two or three runners have a chance.
This guide explains how that works, what it costs, and when it might make sense. It is about the mechanics. Where we mention ideas, we say so.
Should you bet more than one horse in a race?
It can make sense when two or three runners genuinely have a chance in your own view, and the combined price still offers value. Whether it makes sense is a matter of arithmetic, not of feeling braver.
Backing several horses in the same race so that each pays roughly the same if it wins is called dutching. The effect is that you win more often, but you win less each time and you stake more each time. It changes the shape of your results. It does not create value by itself.
We have not tested dutching on our own database for this guide, so we do not quote results for it. What follows is how the sums work, plus ideas for you to test.
Why one-horse betting feels right but can be hard to live with
Many tipster services sell one name per race, and that format can leave people feeling a good bettor delivers a single pick. It is worth questioning, because a service that named five horses per race would have little to sell.
But races do not always work that way. Open a competitive handicap and you will often see two or three horses with a credible chance, a few outsiders, and the rest making up the numbers. Forcing yourself to choose one can mean ignoring your own evidence.
There is a human cost too. Backing one horse at a time usually means a modest strike rate (the share of bets that win), and that means losing runs. In our verified data (Great Britain and Ireland, 1 July 2013 to 14 October 2026, Betfair SP, one point level stakes, after 2% commission), the Betfair SP favourite won 34.4% of 167,257 races and its longest losing run was 28 races. The Race Predictor top horse (our own model) won 25.5% of 167,314 runners, with a longest losing run of 36. Those figures describe that stretch of racing and will not repeat exactly.
Long losing runs are normal for single-horse betting. Many people start tweaking, raising stakes or abandoning a plan part-way through one. If you want to prepare for that, see losing runs.
What dutching actually does to your strike rate
Dutching means staking on more than one horse in a race, sized so that whichever one wins returns the same amount. On Betfair you can work it out with a calculator.
Here is a worked example. It uses round numbers and ignores commission. It is an illustration, not a result.
Suppose you like two horses. One is 2/1 (3.0 in decimal odds) and the other is 3/1 (4.0). You want £60 back if either wins.
- Stake £20 on the first horse (£20 x 3.0 = £60).
- Stake £15 on the second (£15 x 4.0 = £60).
- Total staked is £35. If either wins you get £60 back, a profit of £25. If neither wins you lose the £35.
Now the key sum. The first horse’s price implies it wins one time in three (33.3%), and the second one time in four (25%). Together that is 58.3%. For the pair to make money, one or other of them needs to win more than 58.3% of the time, in your honest view. The pair wins more often than either horse alone, but you pay for that in the prices.
That is why dutching changes the experience more than the value. Your any-win strike rate rises. Each win is smaller. The edge, if you have one, is whatever sits in the gap between your view of the horses and the combined price.
When dutching might pay, and when it might not
Dutching is not a blanket strategy. It cannot create an edge that the underlying selections do not have. If your picks are poor value, spreading across two of them gives you two poor-value bets.
So the useful question is not “should I dutch?” but “does my reading of this race justify two bets?” Some ideas, offered as ideas rather than rules:
- Races where your own work gives two horses a clear chance and the rest of the field little.
- Races where the market price on each of them is longer than your own estimate would call fair.
- Races where one horse has the form and another has the conditions, and you cannot split them.
Races where you have only a loose feel are probably ones to pass. This is a judgement call, and we have not tested these ideas. If you want to test them, how to test a system live shows one way, and calculate your edge explains the sums.
The field-size question
A common instinct is that dutching is for big fields, where any single selection is unlikely to win. It is worth checking that instinct against the arithmetic.
In a big field each price is longer, so the combined price of two horses is often a smaller number than in a small field. In a small field the two shortest horses can add up to a much higher implied chance, and then you need them to win most of the time for the bet to work.
So field size changes how often a dutched bet returns and how much it pays. It does not tell you whether the price is good. The right reason to dutch a big field is not “the favourite is less likely to win”. It is “I genuinely think two horses have a chance and the prices are long enough”.
How to decide when a race is worth dutching
Here is one way to put it into practice. Treat it as a suggested routine and test it before you rely on it.
- Form your own view of each contender’s chance of winning, as a percentage. This is the hard part. Estimating chances is difficult, and dutching cannot make a guessed percentage reliable. It only helps you compare your view with the prices on offer.
- Add the percentages for the horses you would back.
- Work out the combined implied chance from the prices (1 divided by each decimal price, added up). Value of this kind is explained in what value means in horse racing betting.
- Only bet if your combined view is clearly higher than the combined price implies.
Two horses is the usual number. Three can work when the field is competitive, but the combined stake grows, and the combined implied chance climbs with it, so there is less room left for value.
The Betfair exchange is a natural place to do this. Exchange prices can be different from bookmaker prices, and you pay commission on net winnings in a market, so factor that in when you size the stake on each horse. Most dutching calculators will do the stake split for you. For a wider look at exchange against bookmaker, see bookies vs Betfair.
Where this leaves you
Backing more than one horse in a race is not greedy. It is a way to match your bets to a race in which your own view honestly points to more than one runner. The main thing it offers is a smoother experience, with more frequent wins. It does not promise a better return.
If you try it, keep your stakes sensible with bankroll management, and keep a record so you can see whether it is working. Our starting point for building a plan is betting systems starting point.
To see how the sums look on real races, you can open today’s race cards and practise working out combined prices on a few of them. A handful of races is for practice, not proof.
