When to Bet Each-Way: the Number That Decides
Almost every explanation of each-way betting spends its length on what an each-way bet is: two bets, one for the win, one for a place, at a fraction of the odds. That part takes a sentence and you have just read it. The question nobody answers is the one you actually face at the moment of striking the bet — whether ticking the box is a better use of the money than not ticking it. That question has an exact answer, it is one division, and the place terms matter more to it than the price does.
Only half the bet is a decision
Start by separating the two halves, because they are not equally in doubt. The win half is the bet you would have made anyway: same selection, same price, same judgement. Ticking the each-way box does not change it. What ticking the box does is add a second, different bet alongside it — a bet that your selection finishes in one of the paid places, at a price derived from the first one.
That reframing does most of the work on this page. You are not choosing between two versions of one bet. You are deciding whether to add a second bet, and a second bet stands or falls on its own price and its own chance, exactly like the first. The fact that the two arrive on the same slip is an accident of how betting shops used to write them down.
Write your fractional odds as f and the place fraction denominator as k — the 5 in 1/5 odds, the 4 in 1/4 odds. Then:
Place part price = f ÷ k · Break-even strike rate = k ÷ (k + f)
A 16/1 chance in a handicap paying 1/4 odds is a 4/1 place bet, and a 4/1 bet needs to land better than one time in five. If you would not take 4/1 about that horse reaching the first four, the each-way box is not a refinement of your bet. It is a separate bet you have just decided against without noticing. If fractions and decimals are not yet second nature, the odds converter moves between them, and how betting odds work explains the implied-probability step this page leans on throughout.
The break-even table
Below is the place half at a range of prices, under the two fractions that cover most British racing. The strike rate column is the proportion of the time your selection has to reach a paid place for the place bet to break even. No price here is a quote from any operator; they are arithmetic examples, and every figure was recalculated rather than copied.
| Win price | Place price at 1/5 | Break-even at 1/5 | Place price at 1/4 | Break-even at 1/4 |
|---|---|---|---|---|
| 4/1 | 0.8/1 | 55.6% | 1/1 | 50.0% |
| 6/1 | 1.2/1 | 45.5% | 1.5/1 | 40.0% |
| 8/1 | 1.6/1 | 38.5% | 2/1 | 33.3% |
| 10/1 | 2/1 | 33.3% | 2.5/1 | 28.6% |
| 12/1 | 2.4/1 | 29.4% | 3/1 | 25.0% |
| 14/1 | 2.8/1 | 26.3% | 3.5/1 | 22.2% |
| 16/1 | 3.2/1 | 23.8% | 4/1 | 20.0% |
| 20/1 | 4/1 | 20.0% | 5/1 | 16.7% |
| 25/1 | 5/1 | 16.7% | 6.25/1 | 13.8% |
| 33/1 | 6.6/1 | 13.2% | 8.25/1 | 10.8% |
| 50/1 | 10/1 | 9.1% | 12.5/1 | 7.4% |
Two things are worth reading off it. The first is how much the fraction costs: at 12/1 the difference between 1/4 odds and 1/5 odds moves the required strike rate from 25 per cent to 29.4 per cent, which is a bigger change than most punters would accept on the win price without complaint. The second is the direction of travel. Long prices need a lower place strike rate, short prices need a very high one, and at 4/1 with 1/5 odds the place half needs to land more than half the time before it pays for itself. That is the arithmetic behind the oldest piece of advice in racing, and it is worth having the number rather than the advice.
The same answer from the other direction
There is a second way to ask the question, and it is the one a punter with a fixed budget actually faces: for the same total outlay, each-way or a bigger win single? Take £5 each-way, which is £10 across the counter, and set it against a £10 win single on the same horse at 12/1 in a sixteen-runner handicap paying four places at 1/4 odds.
If it wins:the each-way bet returns £85 — £65 from the win half at 12/1 and £20 from the place half at 3/1. The £10 win single returns £130. The each-way punter has given up £45.
If it finishes second, third or fourth: the each-way bet returns £20 and the win single returns nothing.
So the trade is £45 against £20, which means the horse has to be 2.25 times as likely to place without winning as it is to win.
That ratio has a formula of its own — f multiplied by k minus 1, all divided by k plus f — and it is worth noticing that it can never exceed k minus 1 however long the price gets. At 1/4 odds the demand tops out at three times, at 1/5 odds at four times.
Here is the part that makes the whole page simpler than it looks. If you rate your selection’s win chance exactly at the price on offer — neither a steal nor a trap — that ratio condition reduces, with a line of algebra, to precisely the k ÷ (k + f) strike rate from the section above. In the worked example both routes give 25 per cent. They are not two tests that happen to agree; they are one test approached from two sides. The budget comparison only parts company with the strike-rate test when you think the win price itself is wrong, and then it parts company in the obvious direction: if the horse is overpriced to win, put the money on the win; if it is fairly priced to win but unusually reliable at getting round into a place, the each-way half earns its keep. The each-way calculator settles the returns for any stake, price and place fraction, including the awkward cases further down this page.
The par price: when the place half stops needing a good horse
A break-even strike rate on its own is hard to argue with, because most punters have no feel for how often a given horse places. There is a crude but honest benchmark available for nothing: a runner drawn at random from the field places as often as the number of paid places divided by the number of runners. Four places in a twenty-runner handicap is 20 per cent.
Set the break-even rate equal to that random rate and solve, and you get a par price for any race shape: k multiplied by runners minus places, divided by places. Above that price, the place half only needs a below-average place chance to break even. Below it, the place half needs an above-average one — a strong claim to make about a horse the market has already sent off at a long price.
5 to 7 runners, 1/4 odds, 2 places
Par price: 6/1 in a five-runner race, 8/1 in a six, 10/1 in a seven
The shortest fields are the hardest place bets to justify, because two places in a seven-runner race is barely more generous than two in a five and the price you need climbs with every extra runner.
8 or more runners, 1/5 odds, 3 places
Par price: 8.3/1 at eight runners, 11.7/1 at ten, 15/1 at twelve, 21.7/1 at sixteen
The fifth-odds fraction is what makes this shape expensive. A sixteen-runner non-handicap pays three places at 1/5 and demands roughly 22/1 before the place half is asking for a below-average horse.
Handicap, 12 to 15 runners, 1/4 odds, 3 places
Par price: 12/1 at twelve runners, 14.7/1 at fourteen, 16/1 at fifteen
The quarter-odds fraction pulls the par price down by roughly a third against the shape above it. Same three places, materially better bet.
Handicap, 16 or more runners, 1/4 odds, 4 places
Par price: 12/1 at sixteen runners, 16/1 at twenty, 20/1 at twenty-four, 26/1 at thirty
This is the shape the folklore is really about, and the par price rises steeply with the field: in a thirty-runner handicap you need 26/1 before the place half stops needing an above-average horse.
The limit of this test should be stated rather than buried, because it is real: a field is not a lottery and its runners do not have equal chances, so the random-runner rate is a benchmark and not a prediction. What it does reliably is tell you which side of the line you are arguing from. Being asked to believe a 33/1 shot places better than one runner in four is a different conversation from being asked to believe it places better than one in eight, and knowing which conversation you are in takes one division.
Place terms are a price, not a rule
Everything above depends on k and on the number of paid places, so it is worth being exact about where those come from. The familiar table is published by the Tattersalls Committee in its Rules on Betting, under the heading of racecourse each-way terms, and it reads as follows.
| Field | Terms | Places paid |
|---|---|---|
| Fewer than 5 runners | All to win | No place betting |
| 5 to 7 runners | 1/4 odds a place | 1-2 |
| 8 or more runners | 1/5 odds a place | 1-2-3 |
| Handicaps with 12 to 15 runners | 1/4 odds a place | 1-2-3 |
| Handicaps with 16 or more runners | 1/4 odds a place | 1-2-3-4 |
Now the part that matters, and that almost no guide says out loud. The Committee’s own preamble describes its jurisdiction as hearing “any betting disputes arising from the application or interpretation of these Rules on Betting or any other betting dispute emanating from on-course transactions”. These are racecourse terms. When you bet online, they reach you only because your operator has adopted them into its own contract with you — and an adopted term can be varied.
That is not a technicality; it is the reason the market exists in the shape it does. It is why two operators can offer different place terms on the same race, why extra-place promotions are possible at all, and why the only authoritative answer to “how many places does this pay” is the one on the slip in front of you. Place terms are a competitive price, quoted alongside the odds, and a punter who compares odds without comparing places is comparing half the bet. The same distinction between a published rule and an operator’s own contract decides what happens when a horse is withdrawn, which our guide to Rule 4 deductions works through in the same way.
Each-way is not confined to racing, either. Season-long outright markets carry place terms of their own, usually on a shorter list of places, and the same division applies without modification: our guide to outright betting and place terms covers the football version, and the Premier League 2026/27 hub tracks the season those markets are settled against.
Four things that move the answer after you have bet
The sum above is settled at the moment you strike the bet. Four ordinary events can move it afterwards, and the first two are the ones that produce the messages asking why a return is smaller than expected.
A horse is withdrawn and a deduction applies
A Rule 4 deduction reprices the winnings on both halves of an each-way bet, not just the win half, because the place part is a bet at a fraction of the same odds. The sum compounds rather than replacing anything: the place price shrinks by the same proportion as the win price. The deduction is knowable before the race once you know the withdrawn horse and its price.
The field drops below a tier threshold
Place terms are keyed to the number of runners in the race. A sixteen-runner handicap that loses one non-runner is a fifteen-runner handicap, and the shape that paid four places at 1/4 now pays three. Whether your bet keeps the terms it was struck at or moves with the field is a term in your operator rules, and it is the single most expensive thing on this page to assume.
Your selection ties for the last paid place
A dead heat does not shorten the price. It cuts the stake: the place part is settled on a reduced stake in proportion to the number of runners tying for the number of places available, and the remainder is lost. On a photo for fourth in a race paying four places, that is the difference between a full place return and a fraction of one.
An extra place is advertised on the race
Extra-place offers change the arithmetic in your favour by raising the number of paid places while leaving the fraction alone, which lowers the break-even strike rate for the place half. They are commercial terms rather than rules, they are usually restricted to named races and capped stakes, and the qualifying conditions are where the value goes back to the operator. The sum on this page still works: recompute the random-runner rate with the higher place count.
The tie is the one worth rehearsing before it happens, because it is the only one of the four that cuts your stake rather than your price: our guide to dead heat rules sets out the arithmetic, and the dead heat calculator does the sum on a reduced place stake. Any term here you have not met before is defined in the betting glossary.
The check, in four lines
None of this needs a spreadsheet at the moment of betting. It needs one division and one comparison, in this order.
Write down the place price, not the win price
Divide your fractional odds by the place fraction denominator. A 16/1 shot at 1/4 odds is a 4/1 place bet; at 1/5 odds it is a 3.2/1 place bet. Until that number is in front of you, you are not looking at the bet you are about to strike.
Turn it into a strike rate
The place half needs to land more often than one time in one-plus-that-price. At 4/1 that is one in five, twenty per cent. This is the number to argue with, because it is a claim about the horse rather than about the price.
Compare it with the random-runner rate
Paid places divided by runners. Four places in a twenty-runner handicap is twenty per cent. If your break-even rate is above the random rate, the place half needs an above-average horse to break even; if it is below, it needs a below-average one. That is a sieve, not a verdict, and it is the fastest one available.
Then decide the stake, once, before the price moves
An each-way bet is two stakes. The total outlay is double what a punter used to win singles expects, and the commonest mistake here is not a bad bet but an accidental doubling of exposure. Decide the total first and split it, rather than deciding the unit and finding out what the total became.
Frequently asked questions
What is the break-even strike rate for the place part of an each-way bet?
Take the place fraction denominator, call it k, and your fractional odds, call them f. The place part is a bet at f divided by k, and it breaks even when your selection reaches a paid place more often than k divided by k plus f. At 12/1 with 1/4 odds that is 4 divided by 16, which is 25 per cent. At the same price with 1/5 odds it is 5 divided by 17, which is 29.4 per cent. This is the only number that decides whether the each-way half of the bet is worth striking, and it depends on the place terms far more than on the headline price.
Is an each-way bet better than a win single of the same total stake?
It is better exactly when the place part clears its own break-even, and the two questions are the same question. Compared with a win single of double the stake, an each-way bet gives up f multiplied by k minus 1, divided by k, of the winnings when the selection wins, and gains the whole place return when it finishes placed without winning. Setting those against each other gives a required ratio of f times k minus 1, all divided by k plus f. If you rate the win chance exactly at the price on offer, that condition reduces algebraically to the same k divided by k plus f strike rate. The two lenses agree to the last decimal.
Which field sizes make each-way betting worthwhile?
The useful comparison is between the break-even strike rate and the rate a runner picked at random would achieve, which is simply paid places divided by runners. Setting them equal gives a par price of k multiplied by runners minus places, divided by places. In a sixteen-runner handicap paying four places at 1/4 odds, par is 12/1. In a twenty-four-runner handicap it is 20/1, and in a thirty-runner handicap it is 26/1. Above par the place half only needs a below-average place chance to break even. Below par it needs an above-average one, which is a much harder claim about a horse that is already priced long.
Who sets each-way place terms?
Nobody sets them for online betting. The familiar table comes from the Tattersalls Committee Rules on Betting, whose own preamble limits the Committee to disputes emanating from on-course transactions, and it describes racecourse each-way terms. When you bet online those terms reach you because the operator has written them into its own contract with you, which is why place terms differ between operators on the same race and why extra-place offers are possible at all. The operator rules page, not a general table, is the document that governs your bet.
What happens to an each-way bet if a horse is withdrawn?
Two separate things, and they are often confused. A deduction reprices the winnings on both halves of the bet, since the place part is settled at a fraction of the same odds. Separately, the number of runners that actually start can move the race into a different place-terms tier, so a sixteen-runner handicap reduced to fifteen can pay three places instead of four. Whether the terms you were given at the time of the bet survive a reduced field is an operator term rather than a general rule, and it is worth reading before the race rather than after the result.
Sources
The sources below were read on 19 September 2026. The place-terms table was taken verbatim from the Tattersalls Committee rules and cross-checked against a second document published by the same body. Every break-even figure, par price and worked return on this page was calculated independently rather than copied from any source, and no price shown is a quote from any operator.
- Tattersalls Committee, Rules on Betting, racecourse each-way terms and the preamble on the Committee’s jurisdiction — tattersallscommittee.co.uk
- Tattersalls Committee, Rules on Betting consultation paper, carrying the same each-way terms table — tattersallscommittee.co.uk
- Gambling Commission, licence condition 7.1.1, fair and transparent terms and practices — gamblingcommission.gov.uk
18+. The each-way box is the easiest way in betting to double your outlay without deciding to, and the point of the sum on this page is to make that a choice rather than a habit. Decide the total stake before the price starts moving, not after. Support, deposit limits and self-assessment tools are on our responsible gambling page.