Each-Way vs Place-Only in UK Racing: When the Single Place Bet Beats the Hybrid

The £20 question that started this whole piece

A friend of mine, a casual punter who bets twice a year on the big races, asked me a question over a coffee in March. He had £20 in his pocket, he wanted to back a horse he liked at 9/1 in the Grand National, and he couldn’t decide between staking the whole £20 on each-way (£10 win, £10 place) or splitting it into a £20 place-only single. He had a vague sense that the place-only bet might be the better idea, but he didn’t know why, and he hadn’t found anyone willing to walk him through the maths without trying to sell him an account opening.

That conversation is the engine for this entire piece, because his question is the question. Roughly three-quarters of stakes on the Grand National itself are placed each-way — industry estimates put the figure at around 74 to 75 per cent — and a great many of those each-way punters would do better with a straight place-only single. Not all of them. Some of them. The skill is knowing the difference, and the difference comes down to maths that takes about three pages to explain properly and then forty years to internalise.

I have been making this comparison for clients for the better part of a decade, and I think the cleanest way to do it is to set out the two bets side by side, derive the break-even point at which they pay identically, and then look at the actual ranges of prices and race types where one structurally beats the other. That is what the rest of this article does.

The two bets side by side, mechanics laid bare

An each-way bet is, despite its single-line appearance on a betting slip, two bets stapled together. You stake an equal amount on the win and on the place. Your total outlay is double the stake you write next to “each-way,” because the £10 each-way is a £20 outlay. If the horse wins, both halves of the bet pay out. The win side returns at the full win odds. The place side returns at the bookmaker’s fractional place odds — typically one-quarter or one-fifth of the win price, depending on field size — for finishing in the qualifying positions.

A place-only single is one bet. You stake whatever amount you like on the place market alone. If the horse finishes in the qualifying place positions (one, two, three, sometimes four, depending on the band), the bet pays out at the place fraction of the published win odds. If the horse wins, the place-only bet still pays — but only at the place fraction. There is no separate win-side return. If the horse finishes outside the places, the bet loses.

The asymmetry between the two structures is brutally simple. The each-way bet pays more if your horse wins, because you collect on both halves. The place-only bet pays more if your horse merely places, because you can deploy your entire stake on the higher-probability event rather than splitting it across the lower-probability one. Which structure is better for a given race depends entirely on what you think the horse is likely to do. The bigger your edge on winning specifically, the more each-way earns its place; the more you think your horse is a placer rather than a winner, the more place-only takes over.

One mechanical wrinkle that catches people out: with an each-way bet, if the horse wins, the place portion pays at the fractional place odds, not the win odds. If you back a 9/1 horse each-way for £10/£10 and the horse wins, your win-side gets £90 (£10 stake × 9/1) plus the £10 stake back; your place side gets £18 (£10 stake × 9/5, which is one-fifth of 9/1) plus the £10 stake. The total return is £128 on a £20 outlay. That’s the structure: the place portion always settles at the place fraction, even if the horse wins outright.

The break-even maths that nobody wants you to do

Here’s the equation that the casual punter never sees. Imagine your horse has a known true probability of winning the race, call it P(win), and a known true probability of placing somewhere in the qualifying spots, call it P(place). Your edge on each-way versus place-only depends on the ratio of those two probabilities, the bookmaker’s price on the win market, and the place fraction.

The break-even calculation is straightforward once you write it down. The expected return from a £10/£10 each-way bet at win odds W with place fraction f, requiring £20 outlay, is roughly:

EV(each-way) = P(win) × [W + (W × f)] × 10 + (P(place) − P(win)) × (W × f) × 10 − 20

The first term is the gain when the horse wins both sides; the second is the gain when the horse places but doesn’t win; the third is the cost of the £20 outlay. The expected return from a £20 place-only single at the same fraction is:

EV(place-only) = P(place) × (W × f) × 20 − 20

Setting those equal and solving for the price at which the two are indifferent gives you a clean rule of thumb. At one-fifth place fraction in the 8+ band, the each-way bet starts to beat the place-only bet structurally when the win odds are around 9/2 (4.5 decimal) and longer, provided P(win) is meaningfully greater than zero. At one-quarter place fraction in either the 5–7 band or the 16+ handicap band, the cross-over moves to around 11/4 (3.75 decimal). Below those thresholds, the win side of the each-way contributes too little expected value to justify doubling your outlay, and the place-only single starts to outperform.

That cross-over isn’t a hard rule. It depends on your model of P(win) and P(place), which the bookmaker’s price only partially captures. But it is a useful anchor. If your horse is priced at less than 5/2 to win, you should be sceptical that the each-way structure improves your expected value over a straight place-only single. The bookmaker is using the doubled stake to extract value from the punter’s overconfidence in the win side, and you should not give them that gift.

Keith Melrose’s observation about short-priced favourites being “the punting equivalent of a slow puncture” lives in this exact territory. The reason short-priced each-ways drain bankrolls is precisely that the place-side return at one-fifth of a short price barely pays back the place stake, while the win side faces a margin-loaded bookmaker price that already pulls the expected value below zero. Place-only on the same horse at least gives you full deployment of stake against the higher-probability event.

A £20 stake worked out three different ways

Theory only sticks when you stress-test it against numbers you can feel. Let me take a single horse — a 7/1 second favourite in a 12-runner handicap on the all-weather, which is the 8+ band, three places paid at one-fifth — and work £20 through three structures: full each-way, place-only single, and a hybrid worth flagging for completeness.

Structure one: £10 each-way, total outlay £20. If the horse wins, win side returns £70 + £10 = £80; place side returns £14 + £10 = £24 (place fraction 7/1 × 1/5 = 7/5). Total return £104 on £20 staked, net profit £84. If the horse places (second or third), win side loses £10; place side returns £24. Total return £24 on £20 staked, net profit £4. If the horse finishes fourth or worse, total loss £20.

Structure two: £20 place-only single. If the horse wins, place pays at 7/5; £20 × 7/5 = £28 in winnings, plus £20 stake, total £48 return, net profit £28. If the horse places, same outcome — £48 total, £28 profit. Notice the symmetry: a place-only single doesn’t differentiate between a win and a place. If the horse finishes fourth or worse, total loss £20.

Now compare. If the horse wins, each-way pays £104, place-only pays £48. Each-way is £56 ahead. If the horse places second or third, each-way pays £24, place-only pays £48. Place-only is £24 ahead. If the horse loses, both pay zero. The question becomes: which outcome do you think is more likely, and by how much?

If you believe the bookmaker’s price of 7/1 is roughly fair — implying around a 12 to 14 per cent win probability — and you think the horse has, say, a 40 per cent chance of placing, then the expected values work out roughly equal. The each-way’s £56 advantage on a win, weighted by a 13 per cent win probability, gives about £7.30 of expected value above the place-only on win outcomes. The place-only’s £24 advantage on a place-not-win outcome (probability roughly 27 per cent, since 40 per cent place minus 13 per cent win) gives about £6.50 of expected value above the each-way on place-not-win outcomes. The two bets are practically a coin-flip in expected value.

Move the price the wrong way, however, and the picture flips. The same horse at 3/1 — implying perhaps a 22 per cent win probability with maybe 50 per cent place probability — would give the each-way a much thinner advantage on wins (now just £24 on a 22 per cent event, expected value about £5.30) but the place-only a fatter advantage on place-not-win outcomes (£24 advantage on a 28 per cent event, expected value about £6.70). At 3/1, place-only edges ahead of each-way in expected terms. That’s the cross-over point in action.

Caan Berry, who writes carefully about each-way versus place-only, has been one of the few public voices in British racing actually working examples like this. He typically lands on the same conclusion: at short to mid prices, place-only is the structurally better bet. At long prices, each-way recovers the edge. The Cheltenham Festival is the laboratory where this divides the field most cleanly — and I cover the festival-specific dynamics over in our place bet calculator explained piece.

The races and prices where place-only wins outright

If I had to identify a single race profile where place-only consistently outperforms each-way, it would be the short-to-mid-price contender in a competitive 8+ handicap. The horse priced between 5/2 and 9/2 in a 12-to-15-runner handicap is the textbook place-only opportunity, because the win price doesn’t pay enough on the each-way side to justify the doubled stake, while the implied place probability is high enough that fully deploying the stake on the place side genuinely improves expected return.

The same logic applies to the second favourite in a competitive Group race with eight to twelve runners. A horse priced at 9/4 or 3/1 in a Group 2 or Group 3 race almost never warrants an each-way bet by the maths. The win side is dragging your expected value down because the bookmaker’s margin on a short price is unforgiving, and the place side at one-fifth of 3/1 (just 3/5) is a paltry return on the place stake. A place-only single at the same price lets you concentrate the entire stake on the higher-probability place event.

Beyond those obvious cases, place-only earns its place in two more scenarios. The first is when you genuinely have no view on whether the horse wins or places but you believe the horse is “live” — meaning your model has it placing more often than the bookmaker implies, without distinguishing between first and any other top-three finish. The place-only single matches your model exactly. The each-way structure forces you to bet on a sub-question (will it win?) that your model isn’t equipped to answer.

The second is when the place fraction is unusually generous because of a bookmaker promotion. An “extra place” promotion that adds a fifth or sixth place to a big-field handicap, without changing the place fraction, makes place-only more attractive precisely because the extra places improve P(place) without changing W or f. The each-way also benefits, of course, but the doubled stake structure doesn’t capture the upside as fully as a place-only single does.

Keith Melrose’s flag at Cheltenham about enhanced place terms making “each-way multiples particularly attractive” is the mirror image of this, and worth flagging carefully: at the festival, the extras layered onto big-field handicaps shift the balance back towards each-way for multiples (where the win-side compounding through several legs is structurally valuable). For singles at the festival, the place-only argument is even stronger than at a midweek meeting. The two structures aren’t symmetric — multiples and singles behave differently because the win-side compounding in multiples is what creates the headline payouts.

The races and prices where each-way still earns its place

I have spent the last several pages making the case for place-only, and I want to be honest: each-way is the right bet often enough that I would never describe it as obsolete. The structural argument for each-way is straightforward — at long enough odds, the win side of the bet contributes meaningful expected value that place-only cannot replicate. The Grand National, where I’ve watched 25/1 outsiders win and turn a £10 each-way into £370, is the archetype: the win price is so long that the win-side contribution to expected value is genuine, and the place fraction at one-quarter of 25/1 (more than 6/1 on the place side) is meaningful enough to make a place-finish a real consolation.

The rule of thumb I use: at win odds longer than about 9/2 in the 8+ band (one-fifth fraction), and at win odds longer than about 11/4 in the 16+ handicap band (one-quarter fraction), each-way starts to earn its keep. Above 8/1 or 10/1 in either band, each-way structurally beats place-only across reasonable assumptions about P(win) and P(place). The bigger the field, the more this is true for handicaps — partly because P(win) is genuinely lower with more runners but the bookmaker’s price has to incorporate that, and partly because the four-places-at-quarter-odds setting in 16+ handicaps gives the each-way’s place side more pop than the standard 8+ band’s three-at-fifth.

Each-way also earns its place when the bookmaker offers a generous each-way concession that doesn’t apply to place-only — and there are several. Best Odds Guaranteed on the win side of an each-way is a real concession, because if the price drifts before the off and your horse wins at a longer price, the win side of your each-way pays at the drifted price. Place-only singles don’t always qualify for Best Odds Guaranteed; the operator’s exact terms vary. Non-Runner No Bet on ante-post each-ways is similarly more generous than on ante-post place-only singles at some operators.

The third scenario where each-way wins is the festival multiple, as I mentioned. An each-way Yankee or each-way Lucky 15 at Cheltenham, with enhanced place terms layered on, compounds the win-side returns through multiple legs in a way that no place-only multiple can match. The variance is brutal, but the headline payouts are real, and if the entertainment value is part of why you’re betting, an each-way Lucky 15 at the festival is a structurally legitimate bet. Most casual punters who back the National each-way are, frankly, betting for the entertainment value of having a horse to cheer for in two distinct ways (will it win, will it place), and that’s a real form of utility too.

The state of place-only availability on UK markets

Here is the bit of the comparison that doesn’t show up on most forums: the UK market doesn’t make place-only singles equally available across operators. Some UKGC-licensed bookmakers price place-only as a standalone product on every race, every day. Others only offer place-only on flagship meetings and on televised races. A few don’t price place-only as a separate product at all and only allow place-side exposure through the each-way structure.

The economic reason is roughly the same across the market. Each-way is the more profitable structure for the bookmaker per stake unit because the doubled outlay gives the operator two sources of margin. Place-only singles, with their full stake deployed on a single market, give the punter the cleanest exposure and the operator a thinner margin. Operators that compete most aggressively on price tend to offer place-only as a default product; operators that compete on promotions and welcome bonuses sometimes lean towards each-way as the default.

From the punter’s perspective, this matters because the right structure for your bet is the structure that exists on the market you’re betting into. If your operator doesn’t price place-only as a single, you have a choice: open an account with one that does, or accept the each-way structure even when the maths slightly disfavours it. Roughly two-thirds of the £2.6 billion in remote betting Gross Gambling Yield generated annually on the UK market — of which horse racing contributes around £766.7 million in any given year, the second-largest sport after football — flows through operators that offer place-only as a routine product. The other third flows through operators where each-way is the dominant or sole structure for place exposure.

The 2024–25 industry statistics from the Gambling Commission, published in the summer of last year, framed the headline this way: total Gross Gambling Yield across all British gambling rose to £16.8 billion, an increase of 7.3 per cent. Online gambling drove the growth, including remote betting on racing. For our purposes, the takeaway is that the market is large enough and competitive enough that place-only singles are widely available — you just have to know which operators consistently offer them.

A practical test before you commit: check whether the operator’s place market for a given race appears as a separate line on the race card, with its own published odds, distinct from the each-way line. If it does, that’s a place-only single market and you can stake it independently of the win market. If place exposure is only available as part of an each-way ticket, you can’t isolate the place side.

Staking discipline when the place is your only target

Switching from each-way to place-only changes your staking pattern in ways that are easy to underestimate. A £10 each-way is a £20 outlay; a place-only single needs you to decide what stake you want to put on the place market alone. The temptation is to convert an each-way habit straight across — “I’d have done £10 each-way, so I’ll do £20 place-only” — but that doubles your bankroll exposure relative to the win-side risk you were originally taking.

My own rule, refined over years of testing on actual bankrolls, is to size place-only singles at roughly the same as the win-side stake of the each-way equivalent — so £10 each-way becomes £10 place-only, not £20. The reasoning is that the place-only structure is a higher-probability event with a smaller per-bet return, and sizing it at the each-way’s total outlay would over-concentrate exposure. If you genuinely believe the place-only is structurally better — which it often is at short-to-mid prices — then the right way to capture that edge is consistent staking discipline, not a one-off stake size increase.

Record-keeping matters more on place-only than on each-way, because the visible return per winning bet is smaller. A 4/1 place-only single in the 8+ band returns at 4/5 — under even money — and a string of “winning” place-only bets at those returns can feel underwhelming if you’re not tracking the cumulative result against the cumulative stake. I track every place-only single in a spreadsheet alongside the each-way alternative I’d otherwise have placed, and the comparison over six months tells you very quickly whether your selection style is better suited to place-only or to each-way at your usual price range.

The Gambling Commission’s data on active accounts — 24.4 million active accounts in remote betting at the end of the most recent quarterly count, with customer funds totalling roughly £1 billion at the same date — is a reminder that the average bankroll in UK racing is small. Most punters are sub-£100-a-month players, and the variance of an undisciplined each-way habit grinds bankrolls of that size down quickly. Place-only, sized properly, reduces variance and stretches the bankroll across more bets — which, frankly, is the entertainment value most punters are paying for in the first place.

One last point on staking that I’d flag specifically for place-only: the temptation to chase a long-priced place is structural to the bet. A 14/1 place-only single at one-quarter the odds in a 16+ handicap returns at 7/2 on a place finish, which is meaningful, and a string of those bets at small stakes can feel sustainable. It isn’t. The base rate of placing a 14/1 horse is low enough that variance will smash through even disciplined staking over short runs. Set a unit size you can lose forty times in a row without feeling anything, and stake it consistently. The discipline is the strategy.

Frequent questions on the each-way versus place-only comparison

What to take away before your next slip

The honest answer to my friend’s £20 question, when I worked the maths through for him over that coffee, was that place-only on his 9/1 selection in the Grand National was the slightly better bet — but only just, and the margin would have flipped if the horse had been 14/1 or longer. He took the place-only. The horse finished fourth, and because the bookmaker had a six-place enhancement on the day, he collected at one-quarter of 9/1, a 9/4 return on his £20, for a £45 payout. He bought the next coffee.

The takeaway from all of this is the takeaway he had: ask the maths question before you stake. The each-way structure is the default in British racing, embedded in casual betting habits, but the default isn’t always the right answer for your specific horse at its specific price in its specific field. At short prices, place-only is the better bet more often than not. At long prices in big fields, each-way recovers the edge. Somewhere between 9/2 and 11/4, depending on the band, sits the cross-over. Find the cross-over for your usual betting style, and the rest is discipline.