Reading Sales Videos: Evaluating a Horse at a Distance

Evaluating a horse from a sales video is a structured viewing skill: watch once for overall impression, then again with the sound off, gait by gait — the walk first, because it is the least fakeable and the least improvable; the canter second, because careers are decided in it; the trot last, because it is the most improvable gait and the one editing flatters most. On each pass, the viewer applies the same fault checklist judges use (rhythm, tension, contact) reads how much of the picture the rider is producing, and notes what the edit avoids showing, because in a seller-controlled highlight reel the omissions carry as much information as the footage.

For the international buyer this is the single most-used skill in the whole buying process: every remote purchase, and most travelled ones, begins with a screen. This article is the evaluation method itself — how to watch, in what order, and against which criteria. The risk-management architecture around it (what footage to demand, proxy riders, the doubled vetting, contract and payment) is the subject of buying a horse unseen and, for the fully remote purchase, remote buying. The text around the video is decoded separately, in evaluating sales adverts and the advert decoder.

What a sales video is

A sales video is an advertisement produced by the party with the most information and the strongest interest in the sale: the horse’s best moments, from chosen angles, on a chosen day, in an edit assembled from an unknown quantity of footage. None of this is dishonest; it is what an advertisement is. The error is watching it as a sample of the horse rather than as a curated exhibit, in which everything shown was selected — and everything not shown was also selected.

Two consequences follow. First, a sales video answers one question only — is this horse worth more of my time? — and it answers it asymmetrically: it can rule a horse out with confidence, but rule one in only provisionally, pending raw footage and the stages beyond. Second, the edit itself is evidence: what the camera dwells on, what it cuts away from, which rein appears and which does not. The seller has already told the viewer, by omission, where the questions are.

The viewing method

The discipline that separates professional screening from browsing is passes: several viewings, each watching one thing, rather than one viewing forming one impression.

First pass: impression, sound on. Watch the whole video once as the seller intended, and note the honest first reaction — type, presence, general balance, whether the horse matches the written profile at all. This pass exists to be recorded and then distrusted: it is the impression the production was engineered to create, and writing it down stops it quietly grading the later evidence.

Second pass onward: sound off, one gait at a time. Music manipulates the perception of rhythm — an edit cut to a beat makes the trot look more cadenced than it is and buries the irregularities the soundtrack is timed to cover; muted, the footage often visibly shrinks. From here the viewing runs gait by gait, in an order that deliberately inverts the video’s own priorities, following evaluating gaits and movement: edits sell the trot, but careers are decided by the walk, the canter and the hind leg.

The order is: walk first, because it is the least fakeable gait, the least improvable, and the one the edit least wants to dwell on; canter second, because from Prix St Georges upward the tests are decided in it and a poor canter resists training; trot last, because it is the most improvable of the three gaits and the one most flattered by camera work and cutting — the gait, in other words, where the video is working hardest and the viewer should trust it least.

The walk: watch what the edit hides

Find every second of walk in the video; there is often very little, and its brevity is the first note to take. What exists is checked against the criteria of the walk: a clear, evenly spaced four-beat rhythm; visible overtrack; swing through the body and a softly oscillating neck; relaxation on a longer rein if any is shown. The disqualifying observation is any drift of the beats toward synchrony on each side — the lateral walk, which caps the walk marks for the horse’s career and is regarded by trainers as close to permanent once installed.

Video is at its least reliable exactly here: short clips, cuts and frame rates can hide a lateral tendency, and tension on the day can suppress a good walk. The rule is therefore asymmetric: a clearly pure, marching walk on video is meaningful positive evidence; an absent, brief or ambiguous walk is a standing question, answered only by continuous walk footage (loose rein, hard surface) obtained on request.

The canter: the professional’s pass

The canter pass applies the criteria of the canter: three distinct beats with a visible moment of suspension; an uphill tendency, the forehand rising out of the stride rather than diving into it; a hind leg that jumps actively under the body; and self-carriage — the horse managing its own balance through corners and circles rather than leaning on the hand. The downward transitions, if the edit has left any in, are the richest frames in the video: a horse that falls heavily to the forehand coming back from medium canter has answered a question about collection that no amount of expressive trot can reopen.

Video-specific checks: whether canter appears on both reins at comparable length; whether any transition within the gait (collected to medium and back, per the pace variants) survives uncut; and, on a made horse, whether the flying changes are shown in sequence or only as single, separately cut highlights.

The trot: admired last, on purpose

By the time the trot is watched deliberately, the viewer has already seen it — it will have dominated the first pass, because it dominates the edit. The discipline, from the trot and the gait-evaluation guide, is to watch a full circuit looking only at the hind leg and the back, ignoring the front leg entirely. A trot that still looks good under that constraint (hind leg stepping under the mass, topline swinging, the front leg’s gesture the result of what happens behind) is a back-mover’s trot and will develop. Spectacular knee action over a rigid back with the hind legs pushing out behind is decoration, and it is precisely the profile that historically stalls when collection begins.

The strategic point matters more than any single check: the trot is the most improvable of the three gaits, and the market prices it as if the opposite were true. A viewer who has already settled the walk and the canter can afford to enjoy the trot — and to remember that an ordinary, correct, elastic trot on an otherwise excellent horse is closer to an opportunity than a flaw.

The fault checklist on video

Across all passes, the same fault vocabulary that judges use doubles as the screening checklist, because each family has a video-legible signature:

• Frame and contact. Where does the nose live across the whole clip — at or in front of the vertical, or behind it? The default matters, not the moment: every horse tips behind occasionally, so the question is the frame in ordinary working paces when nothing is being asked. A poll carried below a bulge in the mid-neck is the structural signature of a curled frame, and footage of the horse stretching forward and down on a light rein answers the connection question directly — its absence is itself information. Head-tilting, an open mouth or a busy tongue belong to the same family (contact faults) and read clearly on a screen.
• Rhythm. The lateral walk and the four-beat canter, checked as above; and irregular or unlevel steps in any gait, which raise a question that is veterinary before it is anything else. Persistent irregularity on a screen is never a negotiating point; it is a reason to stop or to route the question to the pre-purchase examination.
• Tension. The signs ladder from the tension article: a held, unswinging back; a swishing tail; a rigid or busy mouth; jogging in the walk; anticipation. On video the context question comes first — is the tension proportionate to the setting? — and the comparison questions second: does the horse settle across the session or wind up, and does the walk survive? One electric clip is a day; the same held back in every clip, including quiet home arenas, is a trait.

Sorted the way the fault families sort for buyers: training faults (green balance, momentary tension) are cheap and belong to the production; installed faults (a lateral walk, a habitual short neck) persist and must be priced; ambiguous signs (irregularity, chronic contact trouble) require professional and possibly veterinary eyes before any money moves.

Reading the riding

The horse is only half of what a sales video shows; the other half is how much the rider is doing, and the distinction between a professional making a well-trained horse look easy and a professional holding a difficult one together is among the most valuable reads on the screen.

The tells are in the rider’s body and the quiet moments. A horse in genuine self-carriage travels on a steady, light contact with the rider’s hands quiet and the leg aids invisible; a horse being managed shows constant small corrections — hands that never stop adjusting, legs that never stop pumping, a position that is working rather than sitting. Transitions and the moments between movements are the richest evidence, which is exactly why edits cut them: watch what happens when the rider stops asking, whether the horse waits or runs on, and what the horse does in any walk break the footage contains. Footage of mounting, the first minutes, or the horse standing still exists almost only when it flatters — and its presence is a genuinely good sign.

None of this condemns professional riding; most sale horses are correctly presented by professionals. The question is transferability (how much of the picture belongs to the horse, and how much departs with the rider) and on it video always understates the problem, because the seller’s rider knows the horse. That gap is what the trial ride, or a proxy rider at distance, exists to measure, and the reason temperament cannot be screened from footage alone.

Editing tells

The edit is read as its own layer of evidence. None of the following proves a fault; each marks a place where the seller chose what not to show, and the accumulation sets the agenda for the follow-up request.

A useful summary discipline: after the final pass, write down what the video did not contain (the missing rein, the missing walk, the missing transitions) and treat that list as the first draft of the footage request.

Beyond the sales video

The sales video screens; it never decides. The escalation path runs from cheap evidence to expensive: requested raw footage first, then a live session, then people. The specification of the continuous, unedited clip to request (stable to cool-down, first minutes cold, walk on a hard surface, both reins with transitions left in, the advertised movements ridden in sequence, conformation stills) is set out in buying a horse unseen; a live video call, in which the buyer directs the seller’s rider in real time, is worse footage and better evidence, because nothing can be cut. The staged protocol that sequences these steps against money is in remote buying. One rule carries from there: a seller’s ease or reluctance in producing raw footage of a horse offered at five figures is itself a data point, and the red-flags vocabulary applies throughout.

Notes, scoring and the discipline of comparison

Screening at volume — and serious searches view dozens of videos — fails without written notes: memory smooths every horse toward its trot, and impressions formed to music do not survive comparison. The working method is a fixed template applied identically to every video: each gait noted against its criteria; the fault checklist run and its findings binned as training, installed or ambiguous; the rider’s workload noted; the editing tells listed; the open questions written as literal questions for the seller. Some buyers score gaits on the 0–10 scale used in young-horse classes to force a number; consistency matters more than calibration, and the gait-evaluation guide’s weighting holds on a screen as on the ground — weight the walk and canter, discount the trot, treat any walk not confirmed pure as unresolved. Where the horse has a competition history, the notes are cross-read against the official record and score sheets (reading a score sheet, buying a competition record): a submission collective persistently below the movement average is the paper version of the held back the video showed.

When video evaluation is enough — and when it is not

Video evaluation is genuinely sufficient for one task: building and cutting a shortlist. Gait purity, the fault families, the broad balance picture and the transferability question are all legible from honest footage, and a horse that fails on the screen fails, cheaply, without anyone travelling.

As a basis for purchase, video is only ever one layer. It cannot show feel under the rider or temperament outside the seller’s routine — the two qualities professionals rank highest — and it proves nothing about soundness. Buyers who proceed at distance do so inside the layered structure built for exactly that gap — raw footage and live sessions, an independent proxy rider, the doubled vetting, contract and insurance, per buying a horse unseen and remote buying — and with profiles forgiving enough to absorb a surprise (schoolmaster or young horse). The tighter the required match between horse and rider, the less any screen can carry, and the stronger the argument that the best response to a promising video is not a transfer but a plane ticket.

Sources

• Fédération Équestre Internationale — FEI Dressage Judging Manual (gait definitions, purity and “on the bit”), effective 1 January 2025. https://inside.fei.org/sites/default/files/FEI%20Dressage%20Judging%20Manual%20-%20Effective%201%20January%202025.pdf
• United States Dressage Federation — Pyramid of Training, 2006. https://www.usdf.org/EduDocs/Training/Pyramid_of_Training.pdf
• United States Dressage Federation — Purpose of Tests (USDF Teaching Manual appendix), 2013. https://www.usdf.org/EduDocs/Training/purposeoftests.pdf
• Fédération Équestre Internationale — Dressage Movements 101 (movement faults), 2020. https://www.fei.org/stories/lifestyle/teach-me/dressage-movements-101-advanced-piaffe-passage-canter-pirouette