Do Class 1 favourites make money?

Pascal, the Lab's eternally hopeful punter, has a soft spot for the top table. Show him a wet Monday seller and he shrugs. Put a Group race or a Grade 1 chase in front of him and he leans in. Class 1, he says, is where the best horses in Britain meet, and that changes the nature of the bet. The best trainers in the country have been aiming at these races for months, and the jockeys in the saddle are the ones everybody wants. The form is deep and fully exposed, because these horses have been running against each other in public for seasons. And nobody is saving a horse for another day here, because a Class 1 prize is the day. Strip away the hidden agendas of everyday racing, he reckons, and what's left is the most trustworthy price in the sport. If the favourite is ever fairly priced, it's here.

He has a point, and a better one than most of the theories on this board. Class 1 favourites are usually very good horses with long public records, and the market has plenty to go on when it builds their price. What needs testing is whether the bet pays, which is a separate matter from whether the horse is good.

So we tested it as bluntly as we test everything. Class 1 is the top tier of British racing: the Group races on the Flat, the Graded races over jumps, and the Listed races just below them. We took every Class 1 race in our database, Flat and jumps alike, 1,416 races in all, and backed the favourite in each one at a flat stake, settled at Starting Price and counted honestly. One warning before you read on, placed where you can't miss it: 1,416 races is a small sample, and we'll be straight with you about how much it can and can't prove.

The pull of the Class 1 favourite starts with respect. Racing fans are taught to admire class, and the whole structure of the sport tells them to: the ratings, the race grades, the black type that marks a Group or Listed winner in a sales catalogue. Backing the best horse in the best race feels less like gambling and more like agreeing with the form book. Frank, the form purist on Pascal's Panel, would put it in one line: class tells, and Class 1 is where it tells loudest.

Then there's the sense of fairness. In a big handicap the assessor loads weight onto the better horses to drag them back towards the pack, so the favourite there is fighting the system. Lower down the grades, some runners are having a quiet run before a bigger target, and the punter can never be sure who's trying. Class 1 feels much freer of all that. These are the races the season is planned around, so every horse is there to win, and the best of them should prove it.

The favourite also wins often enough to keep the faith. It wins 35% of these races, about one in three, and when a Class 1 favourite wins it tends to win the way a good horse should, which makes the memory stick. The beaten ones get explained away with the trip, the ground, a slow pace or a rough passage. Every defeat comes with a reason attached, and a reason never feels like evidence against the theory.

So Pascal's theory gets the racing right. The best horse does usually turn up, and the favourite wins a good share of these races. Where it goes wrong is the step from a good horse to a fair price. That step is where the money goes, and it's the subject of the next section.

Everything Pascal likes about Class 1 racing is true for everyone else in the market too. Deep, exposed form is form the whole world has read. A horse with a long public record is a horse that punters, professionals and the bookmakers' own traders have all studied. These are the most scrutinised races in British racing, they draw the sharpest money, and the price on the favourite carries all of it. By the time the stalls open or the tape goes up, there's little a punter knows about the favourite that the market hasn't priced.

That leaves the favourite priced close to its true chance. On its own, a price that close to fair would be a break-even bet at best. Then the bookmaker's margin goes on top. Add up the chances implied by every price in a race and they come to more than a hundred per cent, and the extra is the house's cut, paid on every bet, winner or loser. A fair price minus the margin is a losing price. It loses slowly, a little on every bet.

This is where the theory trips over itself. Quality is not the same as a good price. A brilliant horse at a short price and an ordinary horse at a long one can be equally poor bets, because whether a bet pays is decided by the price. Plenty of money presses on a Class 1 favourite as well. Fans want to be with the best horse, the sharpest judges often agree with them, and the price is cut to the bone before you take it.

Then there's jumping. Class 1 includes the Graded races over hurdles and fences, and a jumps favourite can be travelling best of all and still fall, unseat its rider or be pulled up. Each is a lost stake, exactly like a beaten favourite, and class doesn't make a horse immune to a mistake at the last.

We took the theory at its word and drew the line where racing itself draws it. Every race in our database carries an official class, and Class 1 is the top of that ladder: the Group races on the Flat, the Graded races over jumps and the Listed races beneath them. Our filter picks every race recorded as Class 1, Flat and jumps together. A race with no class recorded in our data never qualifies, since we leave it out instead of guessing. That gives 1,416 races out of the 30,221 real British races we hold.

In every one we backed the favourite: one flat stake, official industry Starting Price, non-runners voided, joint-favourites split, and a recorded winner required. Over jumps, the fallers and pulled-up horses count as the losing bets they are. It's the same honest settlement as every other experiment on this board, with no selection layered on top, no staking trick and no hindsight.

Now the caveat, in the middle of the method where it can't be missed. 1,416 races is a small sample, and it has to be, because Class 1 races are a thin slice of the fixture list. A small sample still tells the truth, but it tells it with a wide margin of error. The 95% range on this result is [-19.3,-5.6], wide enough that we'd be foolish to quote the headline figure as if it were settled. Our everyday favourite test rests on 30,221 races and pins its number much more tightly. This one wears its uncertainty openly, and we want you to see it.

What a sample this size can do is answer the yes-or-no question it was asked. If Class 1 favourites carried a profit, a real edge would be expected to show itself somewhere across 1,416 races. The direction of the answer is clear. How big the loss is, to the nearest point or two, is the part still settling.

Here is the result, and the range deserves as much of your attention as the headline. Backing the favourite in every Class 1 race returns -12.39% across 1,416 races, so about £87.61 of every £100 staked comes back over the long run. The favourite wins 35% of these races, which means Pascal is right that the best horse usually shows up. It just doesn't show up often enough, at the prices on offer, to cover the times it gets beaten.

The 95% range is [-19.3,-5.6]. Read that before the headline figure. The whole range sits below zero, so even the kindest reading of the data shows a loss, and the painful end of it is a heavy one.

For context, the favourite across all 30,221 races returns -9.0%, and the festival favourite, backed in every race worth forty thousand pounds or more, returns -8.2%. Put the three side by side and it's tempting to read a story into the gaps, that the top races are even harder on the favourite backer than ordinary racing. We won't tell that story. The Class 1 range is wide enough to take in the everyday figure, so the honest reading is that the difference sits within the noise. With 1,416 races we can't say whether Class 1 favourites lose more than everyday favourites, less, or about the same. We can say they lose.

That's the finding, and it's a quiet one. The best-studied favourites in British racing, in the races the whole season is built around, follow the same rule as every other favourite on this board. All the depth of form went into the price, and the margin was charged on top of it.

These figures live-pull from the current data, so this is the page most likely to move as seasons pass. Every year adds Class 1 races, the range tightens, and the size of the loss will come into sharper focus.

On this evidence, Class 1 favourites don't make money. Across 1,416 Group, Graded and Listed races the favourite returns -12.4% to Starting Price, and the 95% range of [-19.3,-5.6] stays below zero from one end to the other. We hold the exact figure lightly, because Class 1 races are scarce and the sample is small. Whether the top tier leaks more or less than the everyday favourite is more than 1,416 races can tell us. That it leaks is clear.

Pascal deserves credit for the racing half of his theory. Class 1 is where the best horses meet, the form is as deep as it gets, and the favourite wins about one in three of these races. He went wrong at one step only, and it's the step that costs money: he assumed the most trustworthy horse would come with a trustworthy price. The scrutiny that makes the favourite reliable is the same scrutiny that sets its price. Once the market has had its say and the bookmaker has added the margin, there's nothing left over for the backer.

That lesson travels to every big race you'll watch. Quality is not the same as a good price. The better known the horse, the more people have already backed the same opinion, and the shorter the price you're offered for sharing it. A Class 1 favourite is usually a good horse at a price that already knows it.

None of this takes anything away from Class 1 racing as a spectacle. It's the best the sport has, and a bet can earn its cost as entertainment on a big afternoon when it's sized as entertainment. Past performance is no guide to future returns, our Starting Price settlement with no commission is kinder than what most punters get in practice, and no staking plan turns a negative edge into a positive one. If you bet the big races, use a flat stake you can afford to lose.