Horse Racing Odds Explained: Fractional, Decimal and the Maths Behind Them

Odds Are Just a Price Tag in Disguise

A friend of mine, a stats teacher who had never bet on a horse in his life, asked me last spring why his nephew’s Cheltenham slip showed “5/2” instead of “3.5”. I told him to imagine a price tag in a shop window. Five over two is not a probability, not a ratio of chances, not a percentage. It is what the bookmaker will pay you for every two pounds you stake. Five pounds profit on a winning two-pound bet, plus your two pounds back. Seven pounds total. Done. The maths is a price, not a forecast.

That moment of confusion is the one almost every new British punter has. Fractional notation looks like a maths puzzle because it is written like a maths puzzle. It is not. It is shorthand for a return on stake, and once you see it as a price tag, every other concept in this guide — decimal odds, implied probability, overround, Best Odds Guaranteed — follows neatly from the same thought.

This is a working guide to the numbers. I will show you exactly how to read fractions, how to convert them, how bookmakers build their margin into the market, and where the price actually contains information you can use. The maths is light. The discipline behind it is what matters.

Fractional Odds: The British Default

Walk into any high-street betting shop in Britain, glance at the screens above the counter, and you will see fractional odds. 4/1. 5/2. 11/10. 11/4. These are the British default for a reason that has nothing to do with maths and everything to do with the history of on-course betting. Bookmakers on the rails at Newmarket two centuries ago shouted ratios because ratios were what punters could remember and check at a glance. The fractions survived the shift to digital because punters who grew up with them still ask for “fives” instead of “six point oh”, and the trade kept the format alive.

I still price my own bets in fractions when I am at a course. Decimals are cleaner for spreadsheets. Fractions are cleaner for shouting across a betting ring.

Reading 4/1, 5/2 and 11/10

The rule is simple. The number on the left is your profit. The number on the right is your stake. A horse at 4/1 — pronounced “four to one” — pays four pounds profit for every one pound staked. A two-pound bet on a 4/1 winner returns ten pounds in total: eight pounds profit plus your two-pound stake. A five-pound bet returns twenty-five pounds in total: twenty pounds profit plus your five-pound stake. The fraction tells you only the profit ratio. Your stake always comes back on top.

5/2 — “five to two”, or in betting-shop shorthand “fives” — pays five pounds profit for every two pounds staked. A ten-pound bet at 5/2 returns thirty-five pounds total: twenty-five profit plus ten stake. The two on the bottom is not divided into the five. The fraction is read as a unit ratio. You can simplify if you like. 5/2 is mathematically identical to 10/4 or 50/20, but you will never see those on a board because bookmakers reduce fractions to their lowest readable form.

11/10 is where new punters wobble. The horse is what we call odds-against by the tiniest possible margin. Ten pounds staked returns twenty-one in total: eleven profit plus your ten stake. 11/10 is the smallest unit step above evens, and it is one of the most common prices in British racing for a short-priced favourite that has not quite hardened into odds-on.

Evens and odds-on

Evens — written 1/1 or sometimes EVS on the boards — means a one-pound profit on a one-pound stake. Ten pounds staked, twenty pounds returned. From evens, the market goes in two directions. Above evens are the odds-against prices: 6/5, 5/4, 11/8, 6/4, 13/8, 7/4, 2/1, 9/4, 5/2, 11/4, 3/1 and onwards. Below evens are the odds-on prices, where the bookmaker has decided the horse is more likely than not to win. Odds-on are usually written with “ON” appended: 4/5, 4/6, 1/2 and so on. 1/2 ON means you stake two pounds to win one pound profit. Risk two, win one, get three back including your stake.

The classic odds-on example British racing fans love to quote is Frankel, the unbeaten Cazoo-era champion, who was sent off at 1/10 ON for one of his late starts. A ten-pound bet on a 1/10 ON winner returns eleven pounds. One pound profit. Ten pounds of risk for one pound of profit. The maths sums up everything that is mathematically uncomfortable about backing odds-on horses: the price assumes near certainty, which means there is almost no margin for the bookmaker to give back to you when you win.

Why fractions, historically

The fraction format survives because it tells the punter, in two glances, what the punter cares about most: how much do I win, and how much do I have to risk? Decimal odds combine those into one number, which is cleaner for software but harder to compare for a person standing at a betting ring trying to decide between 5/2 and 11/4. The fractional difference between 5/2 and 11/4 is intuitive. The decimal difference — 3.50 to 3.75 — looks smaller than it is. That intuition matters when you are deciding whether to take a price now or wait for it to drift.

For the same reason, I keep mental shortlists of fractional values rather than decimal ones. 6/4 means win three pounds on two staked. 9/4 means win nine on four. 11/4 means win eleven on four. The pattern repeats up the price list, and once it is internalised the boards in a betting shop start looking less like a foreign language and more like a sentence.

Decimal Odds and Why Some Punters Prefer Them

Decimal odds — 5.00, 3.50, 2.10 — are the European and Australian default and the format every betting exchange uses. The number is your total return per unit staked. Five pounds at 3.50 returns 17.50 in total: 12.50 profit plus your five pounds. The fraction has been folded into one figure, which is why software likes them. Multiply stake by decimal odds and you have your return. No mental arithmetic. No extra step to add the stake back.

Most British online bookmakers now let you toggle between fractional and decimal in the account settings. I keep mine on decimal when I am using exchange tools and on fractional when I am scanning early prices on a typical bookmaker site. Use whichever format makes your decisions faster. Mixing them in the same session is how you end up confusing a 6/4 with a 4.60 and staking twice what you meant to.

Converting fractions to decimals

The conversion is one operation. Divide the top of the fraction by the bottom, then add one. 4/1 becomes 4 plus 1, which is 5.00. 5/2 becomes 2.50 plus 1, which is 3.50. 11/10 becomes 1.10 plus 1, which is 2.10. 1/2 ON becomes 0.50 plus 1, which is 1.50. Add one because decimal odds include your stake in the return. Fractional odds do not.

Going the other way is the same operation in reverse. Decimal 4.50 minus 1 is 3.50, which can be written as 7/2 — three and a half to one. Decimal 1.25 minus 1 is 0.25, which is one-quarter, or 1/4 ON. The maths is trivial. The discipline is doing it consistently. I have a printed conversion card in my desk drawer for the awkward ones — 5/4 is 2.25, 13/8 is 2.625, 11/4 is 3.75 — and I check it every time I have not used a particular price in a while.

Decimal odds and exchanges

Exchanges — Betfair is the dominant one in Britain — display all prices in decimal because they are matching punters against punters, not bookmakers against punters. An exchange is a marketplace. One user offers to back a horse; another offers to lay it. The decimal price is what they meet at. The exchange takes a commission on net winnings — typically two to five per cent — rather than building a margin into the price the way a traditional bookmaker does. As a result, exchange odds are usually a fraction higher than the high-street price on the same horse, especially in the minutes before the off. The Betfair pre-off win market on Boxing Day 2024 turned over £11 216 744 across the British cards, down fourteen per cent on the same day’s £13 031 239 in 2023 — a one-meeting snapshot of how exchange liquidity has shifted year-on-year, and a reminder that even on the biggest betting days the market is not unlimited.

I will treat exchanges in more depth shortly. For now, the point to absorb is that decimal odds are the language exchanges speak, and the modest premium you can find on exchange prices is real, but accessing it costs you commission and a different kind of discipline.

From Odds to Probability

This is where odds become useful as information rather than just a price. Every set of odds implies a probability, and the implied probability is the bookmaker’s estimate of the horse’s chance of winning — adjusted upwards to bake in their margin. Pull that adjustment out and you have a clean view of what the market thinks. Compare it to what you think, and you have a value bet or a pass.

The conversion is straightforward. For decimal odds, divide one by the decimal price. A horse at 5.00 has an implied probability of 0.20, or 20 per cent. A horse at 2.10 has an implied probability of 0.476, or 47.6 per cent. A horse at 1.50 has an implied probability of 0.667, or 66.7 per cent. For fractional odds, divide the bottom by the sum of top and bottom. 4/1 is 1 divided by 5, which is 20 per cent. 5/2 is 2 divided by 7, which is 28.6 per cent. 1/4 ON is 4 divided by 5, which is 80 per cent.

Here is where empirical research becomes useful. The favourite-longshot bias is one of the most-studied anomalies in betting markets. The pattern, observed across the US, the UK and Australia and confirmed by peer-reviewed work going back decades, is that bets on longshots lose, on average, much more than bets on favourites. Studies of UK and Ireland data over a ten-year window have shown that blind backing of favourites produces around a –7.18 per cent return on investment, while blind backing of the ninth or tenth favourite in each race produces losses of forty per cent or worse. Snowberg and Wolfers, writing in the Journal of Political Economy in 2010, argued the bias is best explained by punters systematically misperceiving probability rather than by genuine love of risk. Either way, the practical lesson is the same: outsiders are more wrong than favourites are, in pricing terms, and the market punishes you for not knowing that.

The implied probability gives you a quick way to act on this knowledge. If you think a 10/1 horse — implied probability 9.1 per cent — has a real chance closer to fifteen per cent, you have a value bet. If you think a 10/1 horse has a real chance closer to four per cent, you do not. The maths is the same in both directions; the discipline is being honest about your own estimate.

Overround: The Bookmaker’s Built-in Edge

Here is a question I ask new punters to walk through. Take a six-runner race, look up the prices, convert each to implied probability, and add them. The sum should be one hundred per cent, because exactly one of the six horses will win. It never is. The sum will be somewhere between 105 and 115 per cent in a typical British race, and the extra percentage points are the bookmaker’s margin baked into the prices. That margin is called the overround, or the book percentage. Some traders call it the juice. It is the bookmaker’s edge, and it is the single most important number in commercial betting.

Consider a worked example. Imagine a six-horse race with prices of 2/1, 3/1, 7/2, 6/1, 8/1 and 12/1. Convert each to implied probability: 33.3 per cent, 25 per cent, 22.2 per cent, 14.3 per cent, 11.1 per cent, 7.7 per cent. Add them: 113.6 per cent. The bookmaker has built in a 13.6 per cent margin. If the bookmaker took an equal volume of bets on every horse at those prices and the race was run fairly, the bookmaker would keep 13.6 pence out of every pound bet over the long run.

As Winter and Kukuk put it in their 2007 work on the favourite-longshot bias for Schmalenbach Business Review, “Empirical studies of horse race betting in the U.S., the UK, and Australia have established the so called favorite-longshot bias. Studies find that on average, bets on longshots lose much more than do bets on favorites.” The overround is the mechanical reason — every horse is priced shorter than its true probability, and the longest-priced horses are squeezed proportionally hardest. The 12/1 horse in the example above might have a “true” probability of around eight per cent if the book added up to exactly 100; pricing it to imply 7.7 per cent costs the punter very little in absolute terms but is the most heavily discounted slot in the book.

The smaller the overround, the better the value, all else equal. Big sprint markets at Royal Ascot can trade as low as 105 per cent. Twenty-runner handicap chases on a quiet Wednesday at Kelso can run 120 per cent or higher. Compare the overround across two bookmakers before you place a bet. The differences are often a full percentage point or more, and over a year of regular betting that difference is the difference between losing slowly and losing fast.

Starting Price, Early Price and Best Odds Guaranteed

British racing has a particular wrinkle in how prices are set on the day of the race, and it is the wrinkle that creates the only major bookmaker concession worth chasing. Three terms matter: starting price, early price, and Best Odds Guaranteed, usually abbreviated to BOG.

How SP is set

The starting price — SP — is the official industry price at the moment a race starts. Historically it was settled by a panel of independent reporters watching the prices in the betting ring at the racecourse. Since 2019 the SP has been calculated algorithmically from a wider pool of on-course and off-course bookmakers, with the Starting Price Regulatory Commission overseeing the methodology. The SP is the default price you receive on a bet placed without specifying a price — for example, a phone bet or an early-morning slip that you ticked “SP” on. It is also the price used to settle Tote bets and many promotional bets.

The SP is not a punter-friendly price. It reflects the weight of late money and the bookmakers’ adjustments right up to the off. If a favourite is well backed in the last ten minutes, the SP will be shorter than the morning price. Taking SP is the betting equivalent of buying a flight on the day of travel — you usually pay the peak fare.

Why early prices matter

The early price, sometimes called the show price, is what bookmakers offer in the hours before a race once their prices firm up. Early prices can be available the night before for big festival races, or the morning of the race for ordinary cards. Taking an early price locks in a number; the price may drift longer or shorten in the hours that follow, but your bet is settled at the price you took. For a horse you fancy strongly, taking an early price that you believe is generous is almost always the right call. For a horse you are unsure about, waiting to see the market move can give you a better signal — or a worse price.

BOG explained with worked numbers

Best Odds Guaranteed is the only major bookmaker promotion in Britain that is mathematically free money for the punter, and it exists because the bookmaker is hedging its own competitive position rather than handing you a gift. The deal is this: take an early price on a horse, and if the SP turns out to be longer than the price you took, the bookmaker pays you out at the SP instead. If the SP is shorter, you keep your original early price. You always get whichever is bigger. My deeper look at how BOG actually works walks through the timing, the typical exclusions and the difference between auto-BOG and manual BOG.

A worked example. You take 5/1 — decimal 6.00 — on a ten-pound bet at nine in the morning. By the off, the horse has drifted to 7/1 — decimal 8.00 — at the SP. If the horse wins, BOG settles your ten-pound stake at 7/1 instead of the 5/1 you took. You collect eighty pounds instead of sixty. Alternatively, the horse shortens from 5/1 to 4/1. If it wins, you collect at the 5/1 you originally took, which is now the longer price.

BOG is not active on every bet, with every bookmaker, on every market. It typically applies only to single win and each-way bets on UK and Irish racing placed on the day of the race; ante-post bets, multiples and some big-field handicaps are commonly excluded. Read the terms before you assume it applies. When it does apply, it is the closest thing to a free option you will find in British racing.

Exchange Odds: The Other Market

The exchange model is a separate market sitting alongside traditional bookmakers, and for any punter serious about value it deserves at least a working understanding. An exchange is not a bookmaker. It does not take a position. It matches punters who want to back a horse against punters who want to lay it, and takes a commission on net winnings.

The practical consequence is that exchange prices are typically a fraction higher than the same horse’s price at a traditional bookmaker, especially close to the off when liquidity is highest. A horse priced at 5/1 — decimal 6.00 — in a high-street shop might trade at 6.4 or 6.8 on the exchange in the minutes before the race. Subtract commission and you are still ahead.

The trade-off is that exchanges require more attention. Liquidity dries up on small cards. A market with only a few hundred pounds in it can give wild prices because the matching is thin. Layers — the side of the bet that takes your money if the horse wins — set the prices, and when there are few layers the spread between back and lay prices widens. In big handicaps and at festival meetings, exchange liquidity is deep and prices are competitive. In a midweek handicap at Lingfield, it can be almost non-existent for an outsider.

Commission is the other catch. Two per cent on Betfair UK as a baseline, rising to five per cent for some users depending on activity. Two per cent is small per bet but eats into thin-margin strategies over a season. I treat exchanges as a place to find a better price on horses I have already decided to back at a high-street bookmaker’s stated price, not as a primary betting venue for beginners.

Reading the Price Before the Race Reads You

Every concept in this guide rolls back to one thing: the price contains information. Fractional or decimal, on the high street or on an exchange, at early-morning show or at SP, the number you see has been set by people whose job is to take your money. Sometimes the number is wrong. Most of the time it is not, and the small per-bet edge it carries against you is what keeps an industry of 3 086 licensed gambling operators in business.

The discipline of pricing your own bet — implied probability, overround, your own estimate of the horse’s chance — does not require advanced maths. It requires consistency. Convert every fraction to a probability before you stake. Compare two bookmakers before you click. Take early prices when you have an opinion. Use BOG when it applies. Walk away from prices where the overround is fat and the field is too big to read.

Do that, and you will not win every bet. You will, however, stop losing the bets that are losable before the tape even goes up. That is the maths working for you instead of against you, and that is the entire purpose of understanding the price.

Frequently Asked Questions

Published by the Horseracing Bet Basics team.