Breeders quote hip scores, and buyers rarely know what the figures describe. The schemes assess joint conformation on standardised images and say less about a dog's future comfort than is often assumed.
What is being looked at
A normal hip is a deep socket holding a rounded femoral head closely, with the two surfaces congruent across their contact area.
Dysplastic joints show a shallower socket, a loose fit, or both, which allows movement between the surfaces during ordinary loading.
That movement damages cartilage over time and drives the arthritic change that produces symptoms, which is why the structural assessment is made in the first place.
How the images are produced
Radiographs are taken in a precisely defined position, with the dog sedated or anaesthetised because the required positioning cannot be achieved in a conscious animal.
The dog must be identifiable and above a minimum age, since the joint is still developing before that point and earlier images are not comparable.
Positioning errors change the apparent geometry considerably, which is why images are rejected rather than scored when the standard is not met.
How the scores are constructed
Some schemes assess several separate features of each hip and sum them, producing a figure where lower indicates a better joint and both hips contribute.
Others assign a single grade per hip from a small set of categories. The systems are not directly interchangeable, and a figure means nothing without knowing which scheme produced it.
A score is interpreted against the breed's own distribution rather than in isolation, since typical values differ substantially between breeds.
Why a score is not a diagnosis
The correlation between structure and symptoms is real but far from complete. Dogs with poor scores are sometimes comfortable for life, and dogs with good scores occasionally develop problems.
Body weight, activity pattern during growth and muscle development all influence how a given joint behaves in practice.
A score is therefore a description of a joint at one point in time, useful for breeding decisions and weak as a prediction for an individual animal.
What it is actually for
The purpose is population-level. Selecting breeding animals from the better end of a breed's distribution shifts that distribution across generations.
This works slowly, because hip conformation is influenced by many genes alongside environmental factors during growth.
For a buyer, the useful question is whether both parents were screened under a recognised scheme and how their results compare with the breed average, rather than what a single number was.