Insights into dressage scoring obtained through inertial sensors

Researchers have used inertial sensors fitted to dressage combinations, identifying qualities that high-level judges favor.

Dressage is an Olympic sport based on collaboration between a rider and horse. The relationship is complex, with both partners responsible for distinct parts of the performance.

The rider learns a test pattern consisting of a series of movements that includes geometric figures performed at various gaits and speeds. The rider is responsible for communicating the requirements of the pattern to the horse nonverbally using almost imperceptible signals called “aids”.

During training, the horse learns to interpret and respond willingly to these subtle cues.

A well-trained horse performs the movements willingly without confusion or resistance. As the level of skill increases, horses and riders compete through a sequence of competition levels that requires improved performance of the basic skills combined with the addition of more difficult movements.

Each movement that a horse and rider perform receives a score from one or more judges based on observable criteria.

The performance of the rider, per se, is scored by judges after the test in a single score called “general impression”. It is based on the harmonious presentation of the rider-horse combination, the rider’s position and seat, and the discreet and effective influence of the aids.

It is awarded up to a maximum of 10 points and has a coefficient of two, so effectively, the general impression score receives up to 20 points.

Sarah Jane Hobbs and her fellow researchers, writing in the journal Animals, noted the findings of several published studies that, taken together, show the difficulty in identifying rider performance that can clearly indicate a higher level of skill.

“For a sport involving two somewhat independent athletes performing a complicated pattern of movements, the task of establishing key performance determinants is challenging,” they said.

Holistically, a rider must regulate trunk balance in accordance not only with the gaits and movements being performed, but also maintain the stability of their head in space.

“The strategy used by a rider to achieve this may be gait and horse-dependent, but it must maintain an impression of harmony and must not interrupt the rhythm of the horse.”

The study team set out to investigate holistic, objective measures of trunk and pelvis posture, stability, and coordination that can be used to quantify overall rider performance in dressage.

They did so by comparing the subjective scores of experienced judges with objective data from inertial measurement units to determine which qualities are rewarded by the judges in the collective marks that form the “general impression” score.

The researchers hypothesized that visual information contributing to the judges’ general impression scores for high-level riders include the impression of where the trunk is in space, how well the motion of the pelvis follows the motion of the horse, and the quality of the movement of the horse below the rider.

The study centered on 20 dressage horses, all assessed as sound, and 16 riders. Thirteen of the riders rode one horse in the study, while three rode two.

Two sensors were mounted in standardized positions on each horse and two were mounted on each rider. Each combination performed a standardized dressage test on a 60m by 20m arena, with video footage taken.

Sixteen certified senior dressage judges were recruited and randomly assigned to judge 20 dressage tests on video.

The study centered on the three rider scores that represented the components of the general impression score at the end of the FEI test — each rider’s position and seat; their use of discreet and effective aids to influence the horse; and the overall impression of harmony between the rider and the horse. The judges scored this based on videoed sections in which sensor data had been extracted and analyzed.

These scores were compared with the data from the sensors that measured three-dimensional accelerations and rotations of the horse’s trunk, the rider’s pelvis, and the rider’s trunk.

The researchers found that the score for a horse’s gaits was most heavily influenced by stride frequency, with a slower frequency being favored. The judges’ scores for posture, effectiveness of aids, and harmony with the horse were most strongly influenced by the asymmetries in a rider’s trunk movements, such that higher scores were associated with fewer rider asymmetries.

“We concluded that the scores for the horses’ gaits were higher in horses with slow stride frequencies ridden by riders with more aligned, symmetrical trunks.

“For the judges’ general impression scores, transverse dynamic symmetry of the trunk was the main contributing variable to the riders’ position, harmony with the horse, and effectiveness of aids.”

The importance of this variable in the performance of high-level dressage riders has not been reported previously, they said.

“Pelvis stability, which is well-recognized as a determinant of performance, contributed to scores for the riders’ position and effectiveness of aids.

“Our findings support the old equestrian saying, ‘the rider’s pelvis belongs to the horse, but the shoulders belong to the rider.'”

The study team comprised Hobbs, with the University of Central Lancashire in England; Filipe Manuel Serra Braganca, with Utrecht University in The Netherlands; Marie Rhodin and Elin Hernlund, with the Swedish University of Agricultural Sciences; Mick Peterson, with the University of Kentucky; and Hilary Clayton, with Michigan State University.

Hobbs, S.J.; Serra Braganca, F.M.; Rhodin, M.; Hernlund, E.; Peterson, M.; Clayton, H.M. Evaluating Overall Performance in High-Level Dressage Horse–Rider Combinations by Comparing Measurements from Inertial Sensors with General Impression Scores Awarded by Judges. Animals 2023, 13, 2496. https://doi.org/10.3390/ani13152496

The study, published under a Creative Commons License, can be read here.