Epsom Derby Trends and Statistics: 20 Years of Data for Smarter Bets

Numbers do not care about sentiment. That is the first thing you learn when you spend years pulling apart Derby form, and it is the thing that saves you money more than any tip or insider whisper. I have sat in press rooms at Epsom listening to breathless predictions about the “moral certainty” in the field, and then watched the data – the cold, unromantic data – point to a different horse entirely. The data does not always win. But over twenty years of results, it wins more often than gut feeling.

Aidan O’Brien has won 11 Epsom Derbys. That single statistic tells you more about the modern race than a thousand words of preview analysis. It tells you that the Derby is not a random lottery. It is a race shaped by repeatable patterns: the dominance of certain training operations, the profile of horses that handle Epsom’s unique demands, and the systematic failure of the betting market to price these patterns accurately. This guide pulls together two decades of Derby data and organises it into the trends that actually matter for your betting.

The Favourite’s Track Record: A 30% Strike Rate and What It Means

Here is a question I ask every punter I talk to before the Derby: would you back the favourite in a race where the favourite wins 30% of the time? Most say yes without thinking. Then I ask the follow-up: would you back it at 2/1? The room goes quiet.

Only three of the last ten Derby winners were sent off as the market favourite. That 30% strike rate is below the average for Group 1 flat races in Britain, where favourites typically win around 35-38% of the time. The Derby suppresses the favourite’s chances because of its unique combination of factors: the severe, undulating track at Epsom, the stamina test of a mile and a half on three-year-olds that have mostly raced over shorter distances, and the sheer depth of talent attracted by the biggest purse in British flat racing.

Among the seven non-favourite winners in the last decade, the starting prices ranged wildly. Some won at single-figure odds, closely grouped with the favourite. Others came from deep in the market – 40/1, 25/1, 16/1 – horses the market barely considered. Lambourn won the 2025 running at 13/2, having led the whole way and beaten the 6/4 favourite Delacroix, who trailed home ninth. That result alone captures the Derby’s capacity to humble the market.

What does this mean in practice? First, automatic favourite-backing is a losing strategy over time at the Derby. The prices are too short relative to the strike rate. A favourite at 2/1 needs to win 33% of the time to break even, and the Derby favourite is landing at 30%. Second, the value in the race tends to sit in the 6/1 to 20/1 range, where horses with genuine winning profiles are available at prices that overstate the risk. Third, the favourite’s unreliability makes each-way betting on mid-range runners structurally attractive, because four places in a race where the top of the market is unreliable means place probabilities are distributed more evenly across the competitive portion of the field.

Career Starts, Last-Out Form and the Winner’s Profile

If I could only know two things about a Derby runner before making my bet, I would want to know how many times it had raced and what it did last time out. Those two data points have been more predictive than pedigree, trainer reputation, or morning workout reports over the past two decades.

Twenty-one of the last 23 Derby winners had no more than five career starts before their victory. That is an extraordinary concentration. It tells you that the Derby is a young horse’s race in the truest sense – won by colts that are still improving, still learning, still finding out what they can do. Horses with six, seven, eight starts are not disqualified from winning, but the historical record says they are fighting against the grain. The lightly raced improver, with enough experience to handle the occasion but not so much mileage that his best is behind him, is the Derby archetype.

Why does this pattern hold so consistently? Partly because the best three-year-olds are protected by their trainers. The most talented colts are given just enough racing to educate them – a maiden, a novice or conditions race, then perhaps one trial – before being aimed at the Derby. They arrive at Epsom with scope to improve, and the race itself often draws out a performance level that their form figures do not fully advertise. Overraced three-year-olds, by contrast, have shown more of their hand, and the market has priced them more accurately. There is less hidden upside.

Last-out form reinforces this picture. Sixteen of the last 24 winners won their previous race before the Derby. The connection between a recent victory and Derby success is not about momentum in any mystical sense – it is about fitness and confidence. A horse that won its trial has proven it is sound, in form, and ready to compete at the highest level. A horse that was beaten in its trial might still win the Derby, but the strike rate drops significantly.

The average winner’s rating over the last decade has been 115, and eight of those ten winners carried a rating of 110 or above going into the race. That gives you a performance threshold to work with. A horse rated below 110 is not impossible – the Derby produces surprises – but the data says it needs everything to fall perfectly. A horse rated 110+ with five or fewer starts and a last-out win fits the profile that has produced the overwhelming majority of winners. If your shortlist does not include at least one horse matching this template, you might be fighting the trend.

Trainer and Jockey Patterns That Keep Repeating

The modern Derby has a protagonist, and his name is Aidan O’Brien. Eleven wins from the last 24 runnings. Three consecutive victories from 2023 through 2025. A record that no other trainer in any era has matched at this rate of dominance. When O’Brien talks about the race, the intensity is unmistakable – he has spoken about how from the moment matings are planned at Coolmore through to the training process at Ballydoyle, there is always a huge emphasis on the Derby, and if they had to pick one race to win, it would invariably be this one.

That level of institutional focus produces results. Coolmore breeds specifically for the Derby, investing in sire lines that stamp stamina and class. The Galileo influence – now passed through sons like Frankel, Dubawi’s rivals, and other branches – has reshaped the genetic landscape of the race. O’Brien’s operation does not just enter the Derby; it engineers campaigns around it, routing horses through specific trials, managing their experience, and timing their peak for the first Saturday in June.

For punters, the O’Brien factor creates a strategic dilemma. In any given year, Ballydoyle might saddle three, four, or five runners. Backing all of them is expensive and usually unprofitable unless the winner starts at a big price. The skill is in identifying which O’Brien runner is the stable’s primary hope. Clues include the jockey booking – Ryan Moore on the first choice is the traditional signal – trial form, and the way the stable discusses each horse in the press. O’Brien is a notoriously careful interviewee, but the hierarchy of his entries usually becomes apparent through the trials.

Beyond O’Brien, the Derby’s trainer trends are less concentrated but still informative. Charlie Appleby’s Godolphin operation has produced winners in the last decade. The Gosden yard, now in its John and Thady Gosden iteration, has a strong Classic record. Dermot Weld and Joseph O’Brien have demonstrated that Irish-trained raiders remain competitive. The common thread is that Derby winners come from stables with Classic experience and resources. Boutique operations and first-time Classic trainers occasionally win the race, but the odds favour those who have been there before.

Jockey patterns mirror the trainer picture. Riders with Derby experience – those who have navigated Tattenham Corner under race pressure, who understand the timing of an Epsom finish – perform disproportionately well. A jockey riding in his first Derby has a significantly lower strike rate than one with five or more rides in the race. This does not mean rookies cannot win, but when comparing two runners of similar ability, the one ridden by a seasoned Derby jockey has an edge that the market does not always fully price in.

Gender and Other Factors the Market Overlooks

Every few years, a filly is entered for the Derby and the racing press goes into overdrive. The narrative writes itself: a bold challenge against the colts, a chance to make history. The data tells a different story. Every Derby winner in the last 20 years has been a colt. The last filly even to run in the race was Cape Verdi in 1998. The last filly to win was Fifinella in 1916 – over a century ago.

This is not a trend that requires deep analysis. It is a near-absolute pattern. Fillies are eligible for the Derby but overwhelmingly bypass it in favour of the Oaks, which is run the day before over the same course and distance. The few fillies that do enter the Derby tend to be exceptional on ratings, but the stamina demands and the intensity of a field of colts create conditions that have been hostile to fillies throughout modern history. If a filly is quoted in the Derby ante-post market, treat that price with extreme scepticism unless the form figures are genuinely exceptional.

Geldings present a different case. They are simply ineligible. The Derby is restricted to entire colts and fillies, reflecting its importance to the breeding industry – the race exists, in part, to test potential stallions under maximum pressure. This means any gelding you see in a Derby discussion has been confused with another race, and you can safely ignore it.

Other factors the market tends to underweight include sire line stamina, the impact of ground conditions on lightly raced types, and the significance of a horse’s running style relative to Epsom’s camber. These are not glamorous angles, and they do not generate headlines. But a horse by a proven stamina sire, comfortable on good-to-firm ground, and capable of racing prominently through Tattenham Corner has a materially better profile than one by a speed sire, untested on quick ground, and reliant on a strong late run. The market prices the headline form – trial wins, ratings, trainer reputation. It is slower to price the quieter factors that Epsom specifically rewards.

Epsom Experience: The Stat That Defies Common Sense

You would think that a horse familiar with Epsom’s quirks – the uphill start, the sweeping descent, the off-camber bend at Tattenham Corner, the undulating straight – would have an advantage over a rival arriving at the course for the first time. You would be wrong.

None of the last 12 Derby winners had any previous racing experience at Epsom. Zero. Not one. This is the most counterintuitive stat in Derby analysis, and it catches punters who overvalue course form year after year.

Why does Epsom inexperience correlate with winning? The most likely explanation is selection bias. The best horses do not need to be tested at Epsom in advance. Their trainers are confident enough in their class to let the horse encounter the track for the first time on Derby day itself. Meanwhile, horses that are given a prep run at Epsom – often in the Derby Trial at the course in late April or early May – tend to be slightly lower in the pecking order. Their connections want to confirm the horse handles the track before committing to the main event, which implies a degree of doubt about the horse’s quality or adaptability.

This does not mean a horse with Epsom experience cannot win. It means that when you see a pundit arguing “this horse has a big advantage because he has already run well at the track,” you should treat that argument with caution. The data does not support it. The horse that arrives at Epsom on Derby day as a mystery – whose handling of the course is unknown to the market – is historically the more dangerous proposition.

For bettors, this trend has a practical application: do not downgrade a horse in your assessment simply because it has never raced at Epsom. If anything, treat Epsom inexperience as a neutral or mildly positive factor, provided the horse’s running style and pedigree suggest it will cope with the track’s demands. The truly useful course intelligence comes not from previous race form at Epsom but from how the horse handles left-handed bends and undulating terrain at other courses – information that is available from trial races without needing an Epsom rehearsal.

Turning Historical Patterns into a Pre-Race Checklist

Data without application is trivia. The purpose of compiling these trends is to turn them into a working filter that narrows the Derby field from sixteen runners to a manageable shortlist of three or four genuine contenders. Here is how I do it.

I start by eliminating. Any horse with more than six career starts gets marked down. Any filly gets crossed out unless the form is historically exceptional. Any runner whose last start was a defeat in a weak race drops to the bottom of the list. These filters alone typically cut the field by a third.

Next, I apply the positive criteria. Does the horse have a rating of 110 or above? Did it win its most recent race? Is it trained by a yard with Classic experience, ideally with previous Derby form? Is the jockey an experienced Derby rider? Is the sire line associated with stamina at a mile and a half? Each “yes” adds confidence. A horse that ticks four or five of these boxes is a prime candidate. A horse that ticks two or fewer is a bet against the weight of evidence.

The subtlety is in how you weight each criterion. Not all boxes are equal. A rating of 110+ from a recognised trial like the Dante carries more weight than the same number earned in a weaker race. A last-out win at a mile and a quarter is more relevant than a last-out win over seven furlongs, because the step up to a mile and a half from ten furlongs is more predictable than from sprint distances. I give double weight to the experience-and-form combination – five or fewer starts plus a last-out win – because that intersection has been so consistently powerful in identifying winners.

The final step is to cross-reference the checklist output against the odds. If my filtered shortlist contains a horse priced at 14/1 that ticks four boxes, and the favourite at 3/1 ticks only two, the value case is clear. The market is pricing reputation and media narrative. The data is pricing performance profile and historical precedent.

I also use the checklist in reverse: to identify horses to oppose. A runner that fails three or more criteria but is trading at single-figure odds is a horse the market likes more than the data supports. That knowledge has value beyond simply not backing it – it suggests that the place money in the each-way market is being loaded onto the wrong horses, which in turn means the genuine contenders further down the betting are underpriced.

I do not claim this method produces a winner every year. It does not. The Derby retains the capacity to surprise, and no checklist can account for the random events of a race run at speed over Epsom’s demanding terrain. But over a run of five or ten years, systematically backing horses that fit the historical winner’s profile at odds that underestimate their chance has been the single most profitable approach I have found. The numbers are patient. They do not argue. They simply repeat, year after year, until you listen. For a guide that weaves these trends into a complete betting framework, the full Epsom Derby betting guide puts the data into practice.

Frequently Asked Questions