Is each-way betting on the favourite good value?

Of all the bets a punter reaches for when nerves kick in, the each-way favourite is the comfort blanket. The pitch is simple and it sounds unanswerable. Back the favourite each-way and you get paid even when it only places, so you cannot really lose, you can only lose less often. Pascal puts it the way most punters feel it: the win bit covers the dream, the place bit covers your back, a proper safety net. Half to win, half to place sounds prudent rather than greedy, the responsible way to back a fancied horse.

The claim leans on two things that are genuinely true. The favourite really does land a place a lot of the time, more than three races in five on our data, so the bet keeps handing you small returns and rarely feels like a clean loss. And each-way really is two bets, so when the win half fails the place half can still pay something back. Put those together and it is easy to believe you have hedged yourself, that you have bought insurance against the favourite getting turned over.

So the question this experiment answers is narrow and fair. If you back the favourite each-way in every race, flat stakes, does the safety net actually hold? Does getting paid when the horse merely places keep you out of the red over the long run, or is it just a softer way to arrive at the same loss? We are not asking whether the favourite is a good horse. We are asking whether wrapping it in an each-way bet is good value, and the only honest way to find out is to settle tens of thousands of real races and count what comes back.

The appeal is emotional before it is mathematical, and that is the whole trick. An each-way slip pays out often, so it feels alive. A beaten favourite that plods into second or third still triggers a return, the betting account ticks up, and your brain files it under win even though you have just been handed back less than you staked. That drumbeat of small payouts is deeply reassuring. You are rarely sitting there with nothing, so it never feels like the steady cost it actually is.

There is also a respectability to it. Going each-way sounds like the sensible, grown-up choice, the opposite of a reckless punt. You are not greedily backing the horse to win, you are covering yourself, splitting the risk, behaving responsibly. Bookmakers are perfectly happy to dress the bet up that way, because the language of caution flatters the punter while the maths quietly works against them.

The deepest part of the trap is mistaking how often a bet lands for whether it makes money. These are two completely different things, and the each-way favourite is the cleanest example on the whole page of why. A bet can pay out in the majority of races and still return less than you stake, as long as what it pays back on average is less than what you put in. The frequent place returns are real, but they are small, and they are funded by the times the favourite wins nothing at all, plus the second of your two stakes that you handed over for the privilege. People remember the placed days and forget that they paid double to get them. A bet that pays out a lot and still returns less than you stake is exactly how a steady loss hides in plain sight.

An each-way bet is not one clever bet, it is two ordinary ones settled at the same starting price, and both carry the bookmaker's margin. The win half loses exactly what any favourite-backer loses, because the favourite's short price already has the overround baked in. Backing the favourite to win runs at -8.8% across this data, so the win half is a loser before the place half even gets involved.

The place half is where the value really sits, and it pays less than most punters expect. Bookmakers pay only a fraction of the win odds for a place, usually a fifth, sometimes a quarter. On a short-priced favourite that fraction is tiny. A 2/1 favourite placed at a fifth the odds returns just 2/5, well under half the place stake back as profit. On an odds-on favourite the place return can be less than the place stake itself, so you are paying out to be paid out. You are buying two slices of the same priced-against-you market, and the cheaper slice is the worse value of the two.

Then there is the way racing actually settles. A horse that falls, unseats or is pulled up loses you both halves of the bet, the win stake and, almost always, the place stake too. Over jumps that is not a rare edge case, it happens often, and every one of those is a dead bet on both lines. Our figure counts those non-finishers as the losing bets they are, settled to SP, with non-runners voided and joint-favourites split, which is the honest way to do it.

So both halves pay the overround, the place half is gutted by the fifth-odds fraction, and the non-finishers take the lot. The bet lands often enough to feel like it is working, but landing is not earning. The place returns never quite cover the races the favourite runs nowhere, and the margin compounds the way every house edge does, quietly working against the bank across a season.

We took 27,676 real British races and backed the favourite in every one of them, each-way, at flat stakes. Each-way means two equal stakes, one on the favourite to win and one on it to place, so the total outlay is two units a race, not one. We settled everything to the industry Starting Price, the official odds returned as each race went off, with no early prices, no Best Odds Guaranteed and no exchange commission. That keeps the test clean and, if anything, kind to the bet, because in the real world the prices you actually take tend to be a touch worse.

The place half was settled on standard bookmaker terms, the same fixed fraction of the win odds and the same number of paying places the high-street firms use, scaled to field size the way they scale it. No enhanced or promotional place terms were assumed, because those are short-lived offers, not a system you can run race after race.

The honesty rules are what make the figure trustworthy. Non-runners and withdrawn horses are voided, as they should be. Fallers and pulled-up horses are counted as the losing bets they really are, on both halves of the each-way slip, because a horse that does not complete costs you your win stake and your place stake, and a race only counts if it has a recorded winner. Joint-favourites are handled by splitting the stake across the tied runners, so no result can leak in through a tie. Then we simply added up every pound staked and every pound returned across all 27,676 races, win half and place half together, and expressed the result as return on investment, what you get back for every pound put through over the long run.

Across 27,676 real British races, backing the favourite each-way to Starting Price returns -8.60%. In plain money, for every £100 you stake you get back about £91.40 over the long run, so roughly −8.6p in every £1 is gone. The 95% range on that figure is [-9.8,-7.4], which means the loss is not a fluke of a short sample or a run of bad luck. It is a real, durable, measured drain, and the true number sits firmly in the red across the whole band.

The favourite places 62% of the time, more than three races in five, and that high hit rate is exactly what makes the bet feel safe. It is also exactly what hides the loss. You are getting a return in most races, the account keeps ticking over, and it never feels like it is costing you. But landing a place is not the same as making money. The place returns are small, cut down by the fifth-odds fraction on a short price, and they simply do not add up to enough to cover the win stakes that lose outright and the non-finishers that take both halves.

Here is the comparison that matters most. Backing that same favourite to win, with no each-way at all, returns -8.79%. Going each-way comes in fractionally less bad at -8.60%, because the place half lands often and softens the swings. But you stake double to achieve that, two units a race instead of one, and the place half mostly hands itself back to the bookmaker. You are paying twice as much to lose at almost exactly the same rate. The safety net does not catch you, it just costs you more to feel safe.

And remember this is measured to SP with no commission and no allowance for prices drifting against you, so a real punter taking real prices sees a touch less back still. Not one of the 24 systems we tested on this data makes a profit, and the each-way favourite is no exception.

Each-way the favourite is two losing bets stapled together, and at -8.60% to SP across 27,676 real British races it loses money just like every other backing line we tested. It comes in a whisker less bad than simply backing that same favourite to win, but only because the place half lands often and smooths the ride. You pay double the stake to feel safer, and the place half, gutted by the fifth-odds fraction on a short price, mostly hands itself to the bookmaker rather than protecting you.

The interesting part is what it actually buys you. Going each-way lands more often and smooths the ride, but it returns less than you stake, which is precisely how the cost hides in plain sight. The frequent small place returns are real and they are comforting, but comfort is not the same as value, and a bet that pays out in most races while still returning less than you stake is the textbook shape of a quiet, certain margin working against you. Counted honestly, settled to SP with non-runners voided, fallers and pulled-up horses as the losses they are on both halves of the slip, and every race carrying a recorded winner, it loses −8.6p in the pound.

So treat the each-way favourite as proof of the rule the whole Lab keeps landing on, not as a way round it. How often a bet wins tells you nothing about whether it makes money, because the bookmaker takes the overround on every line you back, and going each-way just gives them two lines instead of one. If you want to see whether anyone holds an edge worth staking on at all, we publish ours either way, losses included, in the track record. This is not a way to beat the bookies.