Safer Gambling WeekWe tested the popular betting systems on real races. See what they madeSet a deposit limit before your next bet. Here is howNo licence means no protection. Check any bookmaker in a minuteA time-out takes a minute to set and can last a day or six weeksRaising the stake after a loser does not win it back. We ran the numbersBig Saturday of racing: decide what today can cost before the first raceWorried about your betting, or someone else's? Free help is a call away

StableBet
Professor Furlong and Pascal at the AI Lab
THE AI LAB
THE LAB · REFERENCE STRATEGIES

Do bumper favourites make money?

No jumps, no fallers, and a yard that knows its horse. Over 761 bumpers the favourite still returns -7.5%.

Doesn't workTested on 761 racesROI: -7.5% ROI
18+ onlyResearch output, not adviceMethodology open · losses visible

Our in-house model lost 16.8% ROI on the pre-registered Oct-Nov 2024 backtest window.

This page publishes what it predicts and tracks every result. We do this because nobody else does. The methodology is open, the losses are visible, the analysis is honest. The model output is presented as a comparison to the market, not as a recommendation to back, lay, or stake on any runner.

Read the full methodology in our in-house AI horse-racing model write-up. Track the running ledger on the Stablebet track record page.

Gambling can be addictive. Please bet responsibly. Free, confidential support from GamCare, GamStop and BeGambleAware. See our responsible-gambling page for more.

The verdict

No. Taking the obstacles away removes the faller, and the market prices the stable's confidence in. Across 761 bumpers the favourite returns -7.5%, in line with favourites everywhere else.

Updated 19 November 2026 · 31,308 races settledSee where this ranks against every system →

What this experiment settles

  • How much does backing the favourite in every bumper cost at Starting Price?
  • If there are no fallers in a bumper, why doesn't the favourite pay?
  • Does the stable's confidence in a well-backed bumper horse give punters an edge?

Methodology

Tested against the Stablebet betting-systems backtest, 31,308 GB races to industry SP, recomputed 19 November 2026. Returns measured to industry SP, flat £10 win on the model's top-rated pick per race unless stated. The underlying ledger and per-race results are public at /ai-race-predictor/track-record/. For the detail, see how the AI model prices a race and how we settle every bet.

By the numbers

−7.5% ROI
−7.5p
Return on every £1
wins 38%
Strike rate
761 bets
Sample
[-16.3,1.3]
95% range

Liam · the stable loyalist swears by this one: the yard and the jockey he follows through thick and thin.

Pascal

Pascal “A bumper is a jumps race with no jumps, so there are no fallers to ruin the bet. Most of the field has little or no form, so when one horse is all the money the yard must know something. Take out the obstacles, follow the stable's confidence.”

Professor Furlong

The Professor It loses at much the same rate as the favourite everywhere else. In bumpers the favourite returned -7.54% to Starting Price over 761 races, about £92.46 back for every £100, and it won 38% of them. Pascal is right that there are no fallers and right that the money carries information. The market can read the same money, so the fancied horse goes off short and the bookmaker's margin is charged on top. The sample is small and its range, [-16.3,1.3], is wide enough to reach about break-even at its kind end, but the central figure is a loss in line with every other favourite. The everyday favourite returns -9.03%.

The claim

Pascal has found a corner of jump racing with no jumps in it. The bumper, a National Hunt Flat race, closes plenty of winter cards: a two-mile gallop for young jumpers who have yet to see a hurdle, run on the same turf as the chases and with nothing to clear. No fallers, no unseated riders, no horse put on the floor by a mistake at the last. Pascal likes that a great deal. His favourites have let him down over obstacles often enough, and here the obstacles are gone.

He has a second reason, and it is a sharper one. Most horses in a bumper have run once or twice, or not at all. The form book is close to blank. So when one of them is all the money, Pascal reasons, the confidence has to be coming from somewhere: the yard has seen it work at home, the owners have paid a lot for it at the sales, somebody knows. Take the jumping out, add a stable that knows its horse, and the favourite should win often enough to pay.

That's a thoughtful theory, and it rests on two things that are true. Bumpers really do remove the biggest single risk in jump racing, and the money in these races really can carry information the public form doesn't show. If backing the favourite pays anywhere on a jumps card, a flat race for unexposed horses is a fair place to look.

So we tested it as plainly as Pascal would bet it. We took every National Hunt Flat race in our database, 761 of them, and backed the favourite in each one with the same flat stake, settled at Starting Price. Then we added it up.

Why people believe it

Start with what Pascal gets right, because there is a lot of it. A bumper takes away the obstacle, and the obstacle is where jump racing hands out its cruellest losers. A favourite travelling well over hurdles can still meet one wrong and come down. In a bumper the worst that usually happens is a horse that runs green, hangs about or gets tired. Every jumps punter has watched a sure thing crash out two from home, and a race where that can't happen feels like the safest bet on the card.

The second half of the theory has history behind it too. Bumpers are where big yards give their young horses a first taste of racing, and they often send out well-bred, expensive horses who have been working nicely at home. Racegoers who follow the major stables talk about "the yard's bumper horse" with real respect. When one of those is the clear favourite, the feeling is that you are getting inside information at a public price.

And the races often look uneven. A field of beginners can include one horse who has already shown a bit and several who have shown nothing, and the market tends to agree loudly about which is which. Watch a hot bumper favourite stride clear up the straight, untroubled by anything, and it is hard to picture that bet losing over a winter.

The trouble sits in the same place it sits for every favourite on this board. Pascal can see the stable's confidence because it is written in the odds. A horse that everyone knows is fancied is a horse with a short price, and the short price is the whole question.

Where the money goes

The stable's confidence is already in the price. Pascal's best argument is that the money in a bumper carries information, and he is right that it does. A well-backed newcomer from a big yard often has something about it. But the money that tells Pascal the horse is fancied is the same money that has shortened its price. By the time he sees a hot favourite, the market has already charged him for the inside knowledge. Being told what the yard thinks is not the same as being paid fairly for it.

Thin form cuts both ways as well. A horse with no public form has no public weaknesses either, so nobody can say for sure that it won't handle the track or won't settle. A favourite in a bumper is often a horse whose reputation runs ahead of anything it has shown on a racecourse. Some live up to it. Plenty of others turn out to be ordinary, and the market only finds out on the day, along with Pascal.

Then there is the bookmaker's margin, the overround. Add up the chances implied by every price in a race and they come to more than a hundred per cent. That surplus is the house cut, and it applies to every runner, favourites included. In a race full of unknowns the bookmakers have every reason to keep the cut healthy, because they are pricing horses they know as little about as anyone.

Taking the jumps away removes the faller. It doesn't remove the price or the margin, and those are what decide whether a bet pays. Young horses also do young-horse things: they run green, hang across the track, get tired or get pulled up, and each of those settles as a losing bet.

Professor Furlong with a losing betting slip at the Stablebet AI Lab
The Professor has run this one through the numbers before. It still loses.

How we tested it

We kept the test as simple as the theory. The filter is the race type recorded in our database: every race listed as a National Hunt Flat race qualifies, and nothing else. Hurdles and chases are out. Flat races on the Flat programme are out too, because a bumper is a jumps-code race for jumps horses, run under National Hunt rules. That leaves 761 races out of the 31,308 real British races we hold.

In each one we backed the favourite: one flat stake, official industry Starting Price, non-runners voided, joint-favourites splitting the stake, and a recorded winner required. A horse that is pulled up, which does happen in a bumper when a young horse runs out of steam or goes lame, counts as a losing bet like any other beaten favourite.

Starting Price is the generous figure. It assumes the official return with no commission and no shaded prices. A punter taking early prices or betting through an exchange would see a different number, and usually a slightly worse one, so if anything this flatters the bet.

There is no stable filter here, no breeding filter and no attempt to pick out the big yards' horses. Pascal's theory is that the money already does that work, so the market's favourite is the test. The sample is the smallest of our jumps tests, because bumpers are a small part of the programme, and the range around the result is wider for it. The next section shows that range rather than hiding it.

The numbers

Here is the number. Backing the favourite in every bumper returns -7.54% across 761 races, so about £92.46 of every £100 staked comes back over the long run. The favourite wins 38% of these races, a touch more often than favourites manage across racing as a whole. The 95% range on the result is [-16.3,1.3].

That range needs reading carefully, because it is wide. Bumpers are a small part of the jumps programme, so this is our smallest jumps sample, and a small sample leaves a lot of room for luck. The kind end of the range sits around break-even and the painful end a long way below it, so this sample on its own can't rule a small edge in or out. What decides it is the central figure: a loss in the same neighbourhood as every other favourite on the board, and those much bigger samples all lose.

The comparisons make that clearer. The everyday favourite across all 31,308 races returns -9.03%. The jumps favourite, hurdles, chases and bumpers together, returns -7.95%, and the favourite over hurdles alone -6.98%. The bumper figure sits right among them. Taking away the obstacles didn't move the return in any direction we can trust, which is the main finding of the page. Pascal removed the faller and the market removed the advantage.

The extra strike rate is worth a sentence too, because it is the part Pascal will notice first. Bumper favourites do win a little more often than the average favourite. The price already knows. A horse that wins more often goes off shorter, and the shorter price takes back what the extra winners give. These figures pull live from our data, so the range will tighten as each jumps season adds more bumpers.

The verdict

Bumper favourites don't make money. Across 761 National Hunt Flat races the favourite returns -7.5%, about £92.46 back for every £100, with a 95% range of [-16.3,1.3]. The range is wide because the sample is small, and its central figure is a loss in line with every other favourite we have tested.

Pascal's theory earns its place on the board all the same. Both of his observations are right. A bumper has no obstacles, so the favourite can't fall at a fence, and the money in these races does carry information about horses the public has barely seen. He read the race correctly.

What he missed is that the market reads it too. The stable's confidence that Pascal wants to follow is the same confidence that shortened the price, so by the time he sees a hot favourite, he is paying for the inside knowledge. The bookmaker's margin sits on top, as it does on every runner in every race, and young horses still find ways to lose that have nothing to do with jumping.

The lesson is the one that keeps coming up on this board. Removing a risk that everyone can see doesn't create an edge, because everyone prices the removal. The favourite in a bumper is a slightly more reliable horse at a slightly shorter price, which leaves the punter roughly where the everyday favourite leaves them.

None of this is a reason not to enjoy a bumper at the end of a winter card. If you bet on one, keep it to a flat stake you can afford to lose. Past performance is no promise of future returns, and our Starting Price settlement is kinder than the prices most punters actually get.

Frequently asked questions

Do bumper favourites make money?
No. Backing the favourite in every National Hunt Flat race in our database, 761 races, returns -7.54% at flat stakes to Starting Price, about £92.46 back for every £100 staked. The favourite wins 38% of those races. The 95% range is [-16.3,1.3]: wide, because bumpers are a small part of the programme, and centred on a loss like every other favourite on our board.
What is a bumper?
A bumper is a National Hunt Flat race: a race on a jumps card with no hurdles or fences. It is run under jumps rules for young horses who are being prepared for a career over obstacles, and many of the runners have little or no racecourse form.
There are no fallers in a bumper, so why doesn't the favourite pay?
Because the market knows there are no fallers too. A favourite that can't fall at a fence is a slightly more reliable horse, and a more reliable horse goes off at a shorter price, so the removed risk is paid for before the bet is struck. The bookmaker's margin then sits on top of the price, as it does on every runner. Young horses still lose in other ways: they run green, hang, tire or get pulled up.
Isn't the money in a bumper inside information?
Often it is, and that is exactly why it doesn't help. The money that tells you the yard fancies a horse is the same money that has shortened its price. By the time a punter sees a hot bumper favourite, the stable's confidence is already in the odds, so following it means paying full price for what the market already knows.
How do bumper favourites compare with other jumps favourites?
They sit right among them. The bumper favourite returns -7.54%, the favourite over hurdles -6.98%, the jumps favourite as a whole -7.95% and the everyday favourite across every race -9.03%. The bumper sample is the smallest of these, so its range is the widest, and we wouldn't read anything into small gaps between them.

What this experiment doesn't cover, and what we're testing next

  • Do bumper favourites from the biggest yards do any better than the rest once the price is counted?
  • Does a bumper favourite's next run over hurdles tell us anything the market missed?

Other Lab experiments

Browse every system tested →