Do favourites win on soft ground?

When the rain arrives and the going stick starts reading soft, a particular kind of confidence arrives with it. Pascal, the Lab's eternally hopeful punter, greets the first heavy card of the winter like a man handed the answers before the exam. Mud, he reckons, is the great sorter. On fast ground anything with a turn of foot can mug you on the line, but soft ground finds out the pretenders: the ones who do not truly stay, the ones with a suspect attitude, the flashy types who hate getting their feet wet. What is left standing, he says, is the proven article. And when the favourite is the proven article, when the formbook shows it has already won in the mud, the race is half over before the tapes go up.

You will hear the same logic everywhere racing is talked about. Ground is the first thing the pundits reach for, the first excuse offered for a beaten odds-on shot, the first filter in every form guide. "Handles the soft, the rest do not" might be the most reassuring sentence in betting. It turns a gamble into homework, and homework feels like it should be paid.

So we tested the homework. We took every British race in our database run on softened going, soft, heavy, and the good-to-soft in between, and backed the favourite in all of them. No cleverness, no exceptions: 10,672 races, one flat stake on the market leader each time, settled at the official Starting Price and counted honestly. If the mud really does sort the field in the favourite's favour, the bottom line will say so. If it does not, we will know exactly what the most reassuring sentence in betting costs to believe, measured to the penny across every wet afternoon in the sample.

The appeal starts with how checkable it is. Most betting theories float on vibes; ground form arrives with receipts. You can pull up a horse's record on soft going in seconds, see the two wins in the Haydock mud, and feel you have done something the crowd has not. A favourite with proven form on the ground does not feel like a punt. It feels like a conclusion.

Then there is the strike rate, the drumbeat that keeps every favourite system alive. The favourite wins 35% of these soft-ground races, roughly one in three, so a backer is constantly collecting. Regular wins feel like a working method, and they quietly hide the only question that matters, which is whether the prices are big enough to pay for the losing days.

The mud adds its own layer of theatre. Soft ground makes races look decisive: fields get strung out, the winner comes clear, the beaten horses are visibly spent. When your mudlark favourite splashes home by six lengths it looks like a law of nature rather than one result. Fast-ground finishes are blanket jobs; heavy-ground finishes look like verdicts, and verdicts are what a believer remembers.

And beneath it all sits the flattering idea that ground study is insider knowledge, the domain of the serious form student. Pascal loves this part most. Everyone can see the favourite, he reckons, but not everyone has checked whether it acts in the mud, so the price must be a step behind the weather. That is the whole theory in one line: rain falls faster than markets move. It sounds shrewd, it feels earned, and it has one problem. The evidence it relies on is printed on the racecard, published in the going report and quoted in every preview, which means the market is not a step behind the weather at all. It read the same forecast you did, hours earlier.

The problem is that nothing about the going is a secret. The official ground description is published before racing and updated through the day. Every past run in the formbook carries the going it was run on. The mudlark's two wins in the Haydock slop are not your discovery, they are a line in every app in the country, and the market has read them too. When the rain comes, the money moves: horses with proven soft-ground form shorten, doubtful actors drift, and by post time the favourite's price already contains everything the weather changed. Backing ground form does not put you ahead of the market. It puts you level with it, at retail prices.

And level with the market is a losing position, because of the engine that grinds down every blind backing line we test: the bookmaker's built-in margin, the overround. Add up the chances implied by every price in a race and they come to more than 100%, about 12% per race extra on a typical field and climbing towards 30% in big fields. That surplus is the house cut, charged on every bet, win or lose, whatever the weather. A market that has already priced the ground correctly and then added its margin leaves the favourite's price a shade too short every single time.

Soft ground even brings its own small taxes. Softened going means more non-completions over jumps, more horses eased once their chance has gone, more late withdrawals as connections dodge the conditions, and none of it surprises the market either. The crowd is right that the mud sorts the field. It is priced-in right. The favourite on soft is very often the correct answer to the question of which horse wins, and still the wrong answer to the question of whether the bet pays. The gap between those two questions is exactly the margin, and the margin does not check the forecast.

The method is deliberately blunt, because the claim is blunt. We hold 26,839 real British races, the whole population, good days and bad. From those we kept every race whose official going description included soft, which captures good-to-soft as well as soft proper, plus everything reported as heavy. That filter is worth pausing on: we used the going as published on the day, not our own after-the-fact reading of the race times, because the published description is the one the punter and the bookmaker both saw when the bet was struck. It leaves 10,672 races, winter jumping and rain-hit summer Flat cards alike.

In every one of them we backed the favourite. One flat stake on the shortest price in the market, settled at the official industry Starting Price. Honest settlement rules throughout, the same ones every experiment on this board uses: non-runners voided, joint-favourites split rather than letting a tie smuggle a winner into the count, a recorded winner required, and over jumps the fallers, unseated riders and pulled-up horses counted as exactly what they are, losing bets. A softened surface produces plenty of those, and a test that quietly dropped them would flatter the bet.

Note what we did not do. No course filter, no trainer angles, no requirement that the favourite had already won in the mud, no distinction between a fashionable mudlark and a horse the market simply liked on the day. The claim we are testing is the general one, that softened ground makes the favourite a stronger bet, so the test is general too. And as always, SP with no commission and no account restrictions is the kind version of reality. A real punter chasing prices in wet weather does a little worse than the number you are about to read.

Here is the number. Backing the favourite in every soft or heavy ground race, flat stakes to Starting Price, returns -8.76% across 10,672 real British races. For every £100 you put through, about £91.24 comes back over the long run. The favourite wins 35% of these races, almost exactly the rate favourites win everywhere, and the 95% range on the sample is [-11.3,-6.1], on the losing side of zero at both ends. This is a steady, durable leak, not a bad run you could wait out.

Now the comparison the whole theory rests on. The favourite across all races and all goings returns -8.6%. The Flat favourite returns -9.1%, the jumps favourite -7.6%. Our soft-ground figure sits in the same narrow band as all of them, and the differences are smaller than the noise in the samples, so we will not insult you by ranking them. The honest reading is that they are about the same. The going changed, the leak did not. Rain redistributes which horses win; it does not redistribute anything from the bookmaker to you.

What the mud emphatically does not do is punish the favourite backer relative to the alternatives. Throw the market's opinion away entirely and pick a horse at random and you lose -21.6%, several times the leak, so the discipline of siding with the market still costs far less than ignoring it. That is the shape of the whole board in one comparison: the market's opinion is excellent and free to borrow, and the margin charged for borrowing it is why the bet still loses. The most reassuring sentence in betting turns out to cost about the same as every other sentence, roughly −8.8p in every pound, delivered with cleaner boots on fast ground and muddier ones in December.

So, do favourites win on soft ground? Yes, constantly, wins 35% of the time, and it makes no difference to the money. Backing them all returns -8.8%, about £91.24 back per £100 staked, the same steady leak the favourite shows on every other surface. The mud is real, the form is real, the sorting effect is real, and all of it is priced before you reach the front of the queue. The weather changes the pictures. It never changes the arithmetic.

That verdict is worth keeping because it generalises. Any angle you can check in ten seconds, the market checked before you. Ground form is the most public piece of private knowledge in racing: printed on the card, recited in every preview, reflected in every price move the moment the forecast turns. Betting on it is not wrong, exactly. It is just not worth anything extra, and the overround turns not-worth-anything-extra into a patient drip of losses.

We would rather you kept the useful half of the idea, because there genuinely is one. Ground study explains races beautifully: why the market moved at noon, why the odds-on shot trailed in, why the winner suddenly looked a different horse in the conditions. As a lens it is excellent, and it will make you a sharper reader of every wet card you ever watch. As a system it returns -8.8%.

The dull, honest rules apply as always. Past performance is not future returns, our SP settlement is kinder than the prices real punters get, and no staking plan can turn a negative edge positive, it only reshapes the swings around the same leak. If you bet in the mud, bet with a flat stake you can afford to lose, or a fraction of the Kelly criterion if you genuinely hold an edge. The going report is not that edge. The market read it first, and charged you for the privilege of agreeing with it.