Methodology

How we score

Version 1.1 · September 2026

Staibles scores every syndicate share we list. This page sets out exactly how, because a rating nobody can check is just an opinion with a number attached. If you think we have got something wrong, tell us — corrections to facts are made promptly, and every listing carries the syndicate's right of reply.

We compare like with like

Before anything is scored, we label what kind of ownership is on offer, because these are genuinely different products and the difference is easy to miss:

• Partnership — a group, usually up to 20 people, who own a percentage of the horse between them.
• Syndicate — members own a share of the horse and of any future sale proceeds, in proportion to their share.
• Lease — members pay to race the horse for a fixed period but acquire no ownership and no entitlement to sale proceeds. The monthly cost can be identical to a purchase share, which is exactly why the distinction is shown on every listing.
• Racing club — members acquire no ownership rights at all. You may share in prize money and enjoy the racing, but you do not own any part of the horse, and there is no resale or residual value.

A £250 club membership, a £250 lease and a £250 syndicate share are not the same thing, and ranking them against each other purely on price would mislead you. So Staibles compares within ownership type by default, shows the type prominently on every listing, and asks you to choose deliberately if you want to look across types.

Two scores, never blended

When you buy a share you are buying two things: a horse, and a deal. A well-bred horse in an overpriced, opaque syndicate is a poor purchase. A modest horse in a fair, open one can be the best fun you will ever have. Blending those into a single number would hide the very thing you need to see, so we publish a Horse Score and an Offer Score side by side.

The Horse Score

Built from public racing data — official ratings, race results, and sire, dam and trainer statistics. Because the evidence base for a horse that has run is completely different from one that has not, we use two weighting profiles.

Horses that have raced

Component	Weight	What it looks at
Form and rating	35%	Official rating, and its direction over recent runs. Trajectory matters more than level — an improving 70 is a better buy than a declining 80
Consistency and activity	20%	Runs per season against the norm for horses of that age and code, and the spacing between them
Trainer factor	15%	Strike rate overall, and with horses of this type, age and class
Pedigree	10%	Sire and dam quality, at reduced weight — the race record now speaks louder
Value against rating	20%	What the share price implies the whole horse is worth, against benchmarks for its rating, age and code

Unraced horses

Component	Weight	What it looks at
Sire quality	30%	Winners to runners, stakes winners, progeny earnings, and — for flat horses — how well the sire's stock go at two
Dam and family	20%	The dam's own record, her produce record, and the quality of close relatives. Where a horse is a first foal, its dam has no produce record yet — that is an absence of evidence, not bad evidence, so we shift the weight to the wider family and say so on the listing
Trainer suitability	20%	Whether this trainer wins with this type of horse. A speed-bred juvenile in a yard that excels with two-year-olds scores higher than the same horse in a staying yard
Purchase quality	20%	What was paid, against sale medians for that sire and sale
Profile fit	10%	Whether the campaign the syndicate describes matches what the pedigree suggests the horse will be suited to

The Offer Score

Built from the terms the syndicate supplies about its own offer, plus public registration records. It is scored independently of the horse.

Component	Weight	What it looks at
Price fairness	22%	What the share price implies the whole horse is worth, against what was actually paid for it. Markup is legitimate — sourcing and management have real costs — but you deserve to see it. Bands are published: under 10%, 10–25%, 25–45%, 45–75%, over 75%. Where a horse was bought at public auction we verify the price against the published sale result by lot number, rather than relying on it being volunteered — so declining to state it gains nothing. Privately bought horses have no public record, and we say so on the listing rather than guessing
Running costs	20%	What a year of the share actually costs, and what that covers. Operators price in incompatible ways — some charge upfront plus a monthly fee, others one all-inclusive payment for a fixed term — so we convert everything to an annual figure before comparing anything. All-inclusive pricing scores higher than a low headline with extras billed separately
Terms and transparency	18%	Prize-money split, term length and what happens at the end of it, what a further term would cost, resale and exit rights, provision for the horse's retirement, and how often you will hear from them. Every item disclosed earns points; every silence loses them
Ownership experience	25%	See below
Operator quality	15%	Whether a BHA licence has been evidenced to us, RSA membership, years operating, horses managed, winners to date as an operator, and whether a complaints process, the insurance position and any contingency fund are published. There is no public register of BHA licences, so where we have not established the position we say exactly that — we never state or imply that an operator is unlicensed. New operators are not penalised to zero — length of trading is one signal among several, not a barrier

Ownership experience

Shares are bought for enjoyment, so scoring only price and terms would measure half the purchase. This component scores only what is published or contractually committed. Marketing language earns nothing — otherwise it would be a competition in copywriting.

Sub-component	Weight	What it looks at
Crowding	35%	Shareholders divided by horses covered. A share in a 40-member syndicate is a different product from one in an 800-member club
Raceday access	35%	How many badges the share you are buying entitles you to, and how they are allocated — guaranteed, a stated minimum, first-come when the horse is declared, first-come with a published priority rule, or balloted. A priority rule, commonly weighted to larger holdings, redistributes access by share size rather than creating more of it, so we score it level with plain first-come and publish the rule beside the share size for you to weigh. Where an operator guarantees a fallback for anyone who misses out — typically a public admission ticket — we credit it, but well below a badge: public admission is not owners' and trainers' facilities, the pre-parade ring or the winner's enclosure. Entitlement often scales with share size, so we score it per listing rather than per operator, and we note whether extra badges can be bought and whether the big festivals are restricted. Where it is not disclosed we do not score this component at all rather than score it zero, and the listing says so
Yard access	20%	Stable visits per year, and whether they are contractual or at the operator's discretion
Communication	8%	A committed cadence of updates — trainer calls, video, written reports — rather than a general promise
Hospitality	2%	Owners' facilities and post-race arrangements

The thresholds, in full

Weightings tell you how much something counts. These tell you how a value becomes a number. They are published because a score you cannot reproduce is not a score you should trust — and because an operator who wants to improve deserves to know exactly what would move it.

Bands are deliberately coarse. A shade under a threshold and a shade over should not produce wildly different scores, and pretending to more precision than the underlying data supports would be false.

Crowding — shareholders per horse

Shareholders per horse	Score
20 or fewer	95
21 to 50	80
51 to 100	62
101 to 300	40
More than 300	22

Where an operator publishes a maximum rather than an actual membership, we use the maximum as a stated worst case and say so on the listing. Supplying the real figure can only help.

Raceday access — how badges are allocated

Allocation	Score
Guaranteed every run	92
First come, first served on declaration	70
First come, with a published priority rule	70
Balloted	55
Purchase only	30

A contractual minimum number of racedays per season adds 8 points. A guaranteed fallback for members who miss out — typically a public admission ticket — adds a further 8, capped at 85 so no combination of additions reaches the value of a genuine guarantee. It is not credited where badges are already guaranteed, since an operator who guarantees never needs one, and it is never counted as an attended raceday: public admission is not owners' access. A priority rule scores level with plain first-come, because it redistributes access by share size rather than creating more of it — so the rule, and the size of the share you are looking at, are published on the listing for you to weigh. Both the entitlement and the allocation method must be published; where either is missing the sub-component is not scored.

Yard access — stable visits

35 points as a base, plus 18 for each visit a year, plus 12 where visits are contractual rather than at the operator's discretion, capped at 100. So one discretionary visit scores 53; four contractual visits reach the cap. The steep early gradient is deliberate: the difference between none and one is the difference that matters most.

Annual cost — against the market median

Cost against the median	Score
30% or more below	92
Below the median	78
Up to 40% above	62
40% to 100% above	45
More than double	30

A broad inclusions list adds 6 points, because fewer unexpected extras is itself worth something. Cost is compared per 1% of the horse owned, not in absolute terms — otherwise an operator selling one-twelfth shares would be marked down for selling more horse. Racing clubs convey no percentage, so they are compared against other clubs on absolute annual cost.

Price against what the horse cost

Implied valuation above the purchase price	Score
Up to 10%	95
10% to 25%	82
25% to 45%	65
45% to 75%	45
More than 75%	25

Only scored where the operator states that the upfront payment buys the share alone. Where the upfront also covers keep already incurred — as many do, and say so plainly — margin and costs cannot be separated, so no markup is calculated and the component is left out. That is a limitation of the arithmetic, not a criticism of the operator.

Overall bands

Offer Score	Band
85 and above	A — Excellent
70 to 84	B — Good
55 to 69	C — Fair
40 to 54	D — Weak
Below 40	E — Poor

A listing where we can compute less than 40% of the framework is not scored at all — a number drawn from one component describes our data gaps rather than the offer. Low-confidence listings are capped at band C until the information is provided.

Year one is not the whole story

Most operators publish one number: what it costs to buy in. But where that price covers the horse's purchase as well as its keep, a second term usually costs considerably less, because the capital is already paid. Other arrangements work the other way, with a low entry price and fees that keep coming. Either way, you often only find out after you have fallen for a horse.

So we score whether the operator tells you, before you buy, what a further term costs and on what basis. Where they publish it, we show year one and year two side by side. Where they don't, the listing says so plainly.

What we will not do is project a three or five year total. Horses leave training through injury, retirement or sale; renewal usually depends on a trainer's report at the end of the term; and many terms are only a year or two long. A confident-looking multi-year figure would be a guess dressed up as a fact, and we would rather show you two real numbers than one invented one.

Cost per attended raceday

The number we think matters most, and one nobody else publishes. There is a difference between the horse running and you being there on an owner's badge — free entry, the owners' and trainers' facilities, the parade ring, and the winner's enclosure if it obliges. The badge is what your share actually buys; anyone can buy a grandstand ticket.

Expected runs come from public data: the horse's age and code, the trainer's typical runs per horse, and the horse's own record where it has one. Badge availability starts from the pool racecourses allocate per runner, multiplied by expected runs and the horses your membership covers, divided by the number of shareholders — then adjusted for how the syndicate distributes them.

We show both figures: cost per expected run, which is about racing frequency, and cost per attended raceday, which is about the experience you actually receive. They can differ enormously, and which one matters is your call. A large club spreading members across ten horses gives you far more runs to follow; a small syndicate with a guaranteed badge gives you more days at the track. Neither is better. They are different products, and you should be able to see which you are buying.

Members of the Racehorse Owners Association can also access racedays through the ROA badge scheme, independently of their syndicate. We mention it on listings, but do not fold it into the calculation — it is a separate purchase you make, and counting it would flatter operators that allocate badges poorly.

What we score, and what we don't

Scores are produced by a fixed set of rules from public racing data and the terms operators supply. They are calculated the same way for every listing. We use AI to explain scores in plain English, to answer questions about them, and to summarise dense terms — never to generate or adjust the numbers themselves.

Some things appear on listings but are deliberately not scored: the yard, the racing colours, photographs, and the names of any notable owners involved. These are what buyers want to see, and they are shown factually — but a recognisable name does not make an offer objectively better, so it earns no points. The same principle applies to the trainer: the name is displayed, but only the strike rate counts.

We do not use veterinary information of any kind, we make no inference about a horse's soundness or health, and we do not predict winners. Racing is uncertain, and no score changes that.

Data confidence

Every score carries a confidence indicator. High means the operator supplied all required information and the racing data is complete. Medium means minor gaps, with benchmarks substituted. Low means material gaps — the score is published with a caveat and capped at the Fair band until the information is provided. This protects you from confident-looking numbers built on thin evidence, and gives operators a straightforward route to a better presentation: tell us more.

Corrections, disputes and changes

• Operators are sent their listing and score before it is published, with a window to tell us about anything factually wrong. We apply corrections ourselves — nobody can edit their own listing, which is what keeps a comparison worth reading. A listing isn't held back by silence, but a correction is welcome at any time, before or after publication.
• Every listing has a right of reply, published alongside the score.
• Factual corrections change scores. Disagreement with the methodology does not — but we would rather hear it than not, and this page changes when the argument is good.
• Weightings are reviewed quarterly against real outcomes, and changes are logged publicly. Each listing keeps a visible score history, so nothing is quietly regraded.

Back to Staibles