How Handicaps Really Work: Inside the Ratings Ladder

Most races run in Britain are handicaps. Set aside the Group contests at the top of the pyramid and the maidens and novices at the bottom, and the everyday sport — the big-field Saturday spectacles, the midweek cards at Catterick and Chepstow, the Grand National itself — rests on one strange premise: that a race can be designed so that every horse in it reaches the line at the same moment.

That is the handicapper's dead-heat dream. Load the best horse with enough weight, ease the burden on the slowest, and in theory their chances equalise; the finish becomes a photograph nobody can split. It has never happened, and it never will, because horses are not machines and opinion is not physics. But the pursuit of that impossible finish is what makes handicaps racing's great puzzle: a standing argument between a small team of officials, who publish their opinions in pounds, and several thousand trainers devoted to proving those opinions wrong.

Almost nobody outside the professional side of the sport understands the machinery. That is a shame, because once you can read the ratings ladder, every handicap in Britain becomes a legible, richly plotted contest.

How a horse gets its number

British handicapping belongs to the British Horseracing Authority, which employs a team of about a dozen handicappers, each responsible for a particular corner of the sport — the official who rates staying chasers is not the one assessing two-year-old sprinters. Each maintains a core group of horses, watches every run, and expresses every judgement in a single currency: the official rating.

A horse must earn its place on the ladder, usually by racing three times so the handicapper has enough evidence to form a view. Winning accelerates the process: a horse successful on debut can be rated after its second start. The opening mark is anchored to the best a horse has shown rather than an average — three quiet efforts worth about 60 will produce a mark close to 60, not a rating built from its worst days.

The Flat scale runs from 0 to about 140, and the top of it is thin air: Frankel's official 140, awarded in 2012, remains the ceiling. Jump racing uses a taller ladder, with the best chasers climbing into the 170s and beyond. The two codes are rated separately, which is why a mark of 150 is fantasy on the Flat and merely excellent over fences.

One rating equals one pound of weight. That is the entire mechanism. A horse rated 90 meets a horse rated 84 in a handicap and carries exactly six pounds more, because six pounds is the handicapper's published estimate of the gap between them. On the card, the top-rated runner takes top weight and every other weight descends from there, pound by pound. Read this way, the weights column stops being administrative furniture and becomes an official opinion of every horse in the race — worth a slower look next time you are working through a racecard.

What a pound buys: lengths and distance

Weight matters because it costs distance, and the exchange rate depends on the trip. The conversions below are approximations — the BHA's own scales flex with the ground and the way a race is run — but the shape is consistent. Over five furlongs on quick ground the official arithmetic prices a beaten length at a little over three pounds; at a mile and six furlongs the same length is worth barely more than one.

• Five and six furlongs: roughly three pounds per length
• Around a mile: roughly two pounds per length
• A mile and a half and beyond: closer to one pound per length
• Over jumps: about a pound per length is the usual shorthand, with far less variation by trip

Turn the ratio around and the consequence appears: a single pound buys a third of a length in a sprint but the best part of a length in a staying race. Weight tells most where horses carry it longest — which is why a rise of five or six pounds can be shrugged off by a sprinter yet quietly end a stayer's season.

Life on the ladder

A rating is never finished. The handicapping team reassesses after every run and publishes revised marks each Tuesday, with changes taking effect for races from the following Saturday. That little gap in the calendar matters, because a horse that wins in the meantime may run again off its old, unrevised mark — usually carrying a fixed penalty, typically six pounds on the Flat and seven over jumps, as a rough stand-in for the rise to come. If the handicapper eventually puts the horse up ten, a six-pound penalty in the interim was a bargain.

Movement on the ladder is asymmetric. Wins bring rises sized to the manner of victory: a scrambling nose in a weak race may cost two or three pounds; six lengths with something in hand invites a far bigger number. Defeat is no guarantee of standing still, either — the official rates the performance, not the placing, so a close second in a strong race can be treated almost as severely as the win. Drops, by contrast, come grudgingly: a pound or two at a time, stretched across a string of defeats. The logic is sound — one bad run proves little, while one good run can prove a lot — but it means a horse can spend months descending a ladder it climbed in an afternoon.

How the numbers are chosen is part science, part craft. The science: race standards, the clock, collateral form — tracing lines through common opponents — and a data analyst on the team. The craft: judging whether the pace was honest, whether the winner idled, whether the margin flatters. Two competent handicappers can defensibly settle two pounds apart, and every Tuesday list is a set of judgement calls wearing the uniform of arithmetic.

The strategic game: staying ahead of the ladder

A perfect handicap would end with twenty horses in a line across the track, and much of the art of training racehorses in Britain consists of making sure it never does.

Because the handicapper can only rate what a horse has shown, the most valuable animal in racing is the one whose ability exceeds its evidence. Racing calls such horses "well handicapped", or "ahead of the mark", and most of the sport's cunning flows from manufacturing that gap — or spotting it in someone else's horse.

The classic case is the unexposed three-year-old in autumn. Its mark was set in spring or summer off a handful of educational runs; by October the horse has grown, strengthened and often moved up in distance, and the number can be a season out of date. Autumn handicaps for that generation are famous among form students precisely because so many runners are still improving faster than their paperwork.

No yard has made the patient version of the art more celebrated than Sir Mark Prescott's Heath House stable in Newmarket. Over a training career of more than fifty years and two thousand winners, the pattern has become a signature: horses begin in modest company, often over trips short of their best, so their opening marks stay low; they step up in distance and improve steeply; and when one comes good it is campaigned at speed, running again within days under a penalty, ahead of the revised rating, sometimes stringing several wins into a few weeks. The most famous product of the method, Pasternak, sealed a four-race winning run in the 1997 Cambridgeshire — the culmination of a plan laid months in advance.

The endgame explains a phrase you will hear on any racecourse: the same horse can be feared off 85 and friendless off 95. Ten pounds is ten rungs. If a horse was eight pounds ahead of the old number, the new one has swallowed the advantage whole — the animal is no worse, but the puzzle has been solved. Handicap specialists consequently care less about how good a horse is than about the relationship between its ability and its mark.

Class bands, age and the fine print

Ratings govern eligibility as well as weight. British races are grouped into seven classes, Class 1 at the top, and most handicaps carry a ratings band in their title: a "0-70" handicap is open only to horses rated 70 or below, a "0-100" to horses rated up to 100, with Class 2 bands stretching into the 100s and Class 6 covering roughly the mid-40s to mid-60s. As a horse's mark rises it is squeezed upwards into stronger races — the ladder again, expressed as the race programme itself.

Age adds a wrinkle. When three-year-olds meet their elders they receive a weight-for-age allowance — a scale introduced by Admiral Rous, appointed the Jockey Club's official handicapper in 1855, and still administered fortnight by fortnight. The allowance shrinks as the season progresses and grows with distance, since immaturity costs most where stamina matters most: by early October a three-year-old receives nothing from older horses over five furlongs yet still takes 11lb over two and a half miles.

Riders can trim the arithmetic too. Apprentice jockeys on the Flat "claim" an allowance while they learn — seven pounds until they have ridden 20 winners, five until 50, three until 95 — shown in brackets after their name on the card, and jump racing's conditional riders work a similar sliding scale. Booking a capable claimer for a top weight is one of the oldest moves in the handicap playbook: the rating stays the same; the saddle simply gets lighter.

The great handicap institutions

Several handicaps are monuments in their own right. The Stewards' Cup, first run over Goodwood's six furlongs in 1840, is a full-field sprint stampede at the height of summer. The Ebor at York, a mile and six furlongs in August, is Europe's most valuable Flat handicap — briefly worth £1 million in 2019, the first British handicap to carry that figure. The Cambridgeshire, run since 1839 over the Rowley Mile's awkward mile and a furlong, forms the historic Autumn Double with the marathon Cesarewitch two weeks later. If you enjoy the way these races anchor the season, the summer festival trail from Glorious Goodwood to the Ebor is in large part a story of handicap plans ripening on schedule.

Above them all sits the Grand National: a maximum field of 34, around four and a quarter miles, and a handicap framed like no other race in Britain. Its weights are unveiled at a ceremony in February, and it is the one race where the handicapper may deliberately step away from strict official ratings — building in proven jumping and staying ability at Aintree, and compressing the top of the handicap, with top weight capped at 11st 12lb, so high-class horses are no longer buried under the crushing burdens of the race's older history. The ultimate handicap, in other words, is also the one where the dead-heat dream is chased hardest.

Reading a handicap with data

For the form student, the official rating is one opinion — authoritative, transparent, and backward-looking by design. The productive habit is to lay other measurements alongside it. Speed figures are the natural partner: where a horse's recent figures sit consistently above its official mark, the ladder may not have caught up with the horse; where they sit below, a supposedly "well-in" runner may be nothing of the sort. Our guide to speed figures explains how time-based ratings are built; Sportily Speed Figures are adjusted for each day's track conditions precisely so that comparison is fair.

Then separate the long-term mark from the recent animal. A horse rated 78 that last won off 84 and has been sliding for a year is a different proposition from a 78-rated improver with three career starts: the first is descending the ladder in instalments, the second may be climbing faster than the handicapper can publish. The rating alone cannot tell you which is which; the shape of the recent form can.

Finally, notice what a model does differently. A modelled win probability re-prices the race from scratch, treating weight as one variable among many alongside pace, trip, ground and class moves. When the model's number and the handicapper's number disagree sharply, one of them is wrong — and working out which is the most interesting question in racing. If you would rather test your theories than trust them, Race Lens lets you build an angle — lightly raced three-year-olds in autumn handicaps, say — from dozens of filters and run it against years of history first.

Sportily is a statistical research platform for racing fans. It does not provide betting advice, and past results never guarantee future performance.