Alignment between gait parameters and judging of canter quality in international vaulting horses

Background: Like most aesthetic sports, judging performance in equestrian vaulting is subjective. The complexity of judging often exceeds human cognitive capacities and can give rise to bias. The horse score makes up 25% of the total score, including the quality of the canter and training level of the horse (A1 score), the execution of the performance and the horse’s willingness and behaviour (A2 score). Objectives: The aim of this study was to compare A1 and A2 scores from judges with objective measurements. Study design: cross-sectional study design Methods: The study explored the alignment between the 6 scores of A1, the A2 score, and 10 measured locomotor parameters in 33 international level vaulting horses. The horses were equipped with an inertial measurement unit (128 Hz) fixed to the sternum to calculate positions and movements and to calculate a quality index. Five international judges scored the relaxation, connection, straightness, impulsion, rhythm, collection and vaultability from the videos in random order. Results: A horse with at least one good score had only good scores and each note was correlated with the same locomotor parameters. A canter considered to be good by the judges was one with good elevation (14 ± 1 vs 13 ± 2 cm; 0.013), good propulsion (6.7 ± 1.4 vs 5.7 ±1.7 m/s 2 ; 0.005) and good regularity of the three beats (0.016 ± 0.004 vs 0.019 ± 0.004 m/s 3 ; 0.018), with the horse slightly inclined to the inside of the circle, balanced on their hindlimbs and with a light forehand. Main limitations: Use of videos for judging Conclusions: These observables seemed to be decisive in assessing the quality of the canter and are of interest in terms of decision support in the context of initial and ongoing training for judges and for considering continuous optimisation of the scoring system. Alignment between gait parameters and judging of canter quality in international vaulting horses

horse, judge, score, kinematic, locomotion, canter. Summary

Like most aesthetic sports, judging performance in equestrian vaulting is subjective. The complexity of judging often exceeds human cognitive capacities and can give rise to bias. The horse score makes up 25% of the total score, including the quality of the canter and training level of the horse (A1 score), the execution of the performance and the horse’s willingness and behaviour (A2 score).

The aim of this study was to compare A1 and A2 scores from judges with objective measurements. Study design: cross-sectional study design

The study explored the alignment between the 6 scores of A1, the A2 score, and 10 measured locomotor parameters in 33 international level vaulting horses. The horses were equipped with an inertial measurement unit (128 Hz) fixed to the sternum to calculate positions and movements and to calculate a quality index. Five international judges scored the relaxation, connection, straightness, impulsion, rhythm, collection and vaultability from the videos in random order.

A horse with at least one good score had only good scores and each note was correlated with the same locomotor parameters. A canter considered to be good by the judges was one with good elevation (14 ± 1 vs 13 ± 2 cm; 0.013), good propulsion (6.7 ± 1.4 vs 5.7 ±1.7 m/s²; 0.005) and good regularity of the three beats (0.016 ± 0.004 vs 0.019 ± 0.004 m/s 3 ; 0.018), with the horse slightly inclined to the inside of the circle, balanced on their hindlimbs and with a light forehand. Main limitations: Use of videos for judging

These observables seemed to be decisive in assessing the quality of the canter and are of interest in terms of decision support in the context of initial and ongoing training for judges and for considering continuous optimisation of the scoring system.

Vaulting is one of the seven equestrian disciplines recognized by the Federation Equestre Internationale (FEI). This discipline consists of gymnastic elements performed on a horse in canter. During a FEI Vaulting Competition, one judge evaluates the performance of the horse (Judge A) including the general quality of the gait and training level of the horse (A1, 60%) by scoring the rhythm, relaxation, connection, impulsion, straightness, and collection; the vault-ability (A2, 25%) by scoring the execution of the performance and the horse’s willingness and behaviour; as well as the quality of the lunging (A3, 15%). The total horse score is 25% of the final score of the test (1). The canter is an asymmetric gait with a synchronisation of the movements of the diagonal pair of limbs separated by a suspension phase. A good canter has a clear beat with a clear moment of suspension. On a circle line, centripetal forces will affect the horse and the horse will lean towards the inside of the circle (inclination), the hooves will be pushing outwards on the ground, and the resultant force moves towards the inside. It is mainly the front limbs that do this work, while the forces of the hind limbs provide propulsion (Hilary M. Clayton et Hobbs 2017). It is expected that the horse shows a degree of collection to carry the vaulter. Faults are sometimes difficult to identify with the naked eye, and we also know that judging aesthetic disciplines is a very complex task, which can exceed human cognitive capacities (Wolframm 2023). Over the past few years, there has been an increasing number of scientific studies into the question of sports refereeing. Studies have looked at objective measures of horse performance (Biau et Barrey 2004), but less on the quality of the canter, and as far as we know, none have studied the quality of the canter on a circle. A few studies have investigated the link between the kinematic performance of the horse and the judges’ scores ((Lewczuk 2013). Consequently, it would be interesting to identify any correlations between kinematic measurements of the vaulting horse and judges’ scores. The aim of this study was to identify objective indicators to help judge canter quality. We hypothesized that propulsion, balance, and inclination are indicators of vaulting canter quality. To this end, 33 international-level vaulting horses were evaluated by five international judges, and their scores were compared to measured locomotor parameters of canter and circle gaits obtained from an inertial measuring unit (IMU) fixed to the horse’s sternum during the test with the vaulter. The current scoring system was analysed and discussed.

Animals A total of thirty-three high-level vaulting horses (13 ± 4 years old with 4 ± 3 years international experience) were included in the present study. All horses had previously competed at international level. There were 29 geldings, two mares and two stallions. They were lunged by 19 lungers familiar to the horses, with 9 ± 4 years of international experience. Study design The study was carried out between 12–2021 and 02–2023 as a multicentre study at six different test centres in four countries (France, the Netherlands, Denmark and Finland). The test area was a 22-meter circle enclosed by a low fence or other suitable material with the lunger in the centre. A vaulter performed a series of predetermined exercises in the following order: Mount on, Basic seat, Stand, Jump from forward stand, Forward swing, Backward swing, Ground jump to the inside, Ground jump to the outside, Sideways prince on the neck, Forward kneeling on the croup, and Handstand in the grips. All tests were recorded with a camera (Go pro Hero 8, GoPro, Inc. 3025 Clearview Way, San Mateo, CA 94402,USA) attached to the front of the lunger. Another camera (Sony HDR CX625, Sony Europe B.V., The Heights, Brooklands, Weybridge, Surrey, KT13 0XW, United Kingdom) was installed outside the circle, in a position that simulated the position of the judge in a competition. This camera filmed each performance from the mount on of the vaulter untill the completion of the final exercise. The horses wore their regular equipment used for vaulting. They were also fitted with an inertial measuring unit (IMU) (Opal sensor, 6g, APDM Wearable Technologies Inc. 7204 SW Durham Road, Suite 800 (Bldg. Q), Portland, OR 97224, USA) attached to the surcingle against the sternum. The sensor included a 3-axis accelerometer, a 3-axis gyro, and a 3-axis magnetometer. The sampling frequency was 128 Hz, and the accelerometers were configured in a high 6G mode. The horses were warmed up before entering the arena for the test. Before starting the recording, the horse/lunger combination had one minute to acclimatize to the arena and side rein type. All horses were recorded with two types of side reins in a randomized design. The side reins join the bit ring and the surcingle. They were an all-leather side rein ”Without elastic”; made entirely of leather, and side reins with a small elastic part ”With elastic” made of leather with an elastic piece measuring 6 cm. The two rein types are both permitted in competition. Between each rein type, the horses’ mouths were checked for lesions, and the horses were hand walked for five minutes without side reins. The side reins were adjusted in accordance with the FEI’s Vaulting Rules (“With the side reins properly adjusted, the horse can carry its nose correctly on or slightly in front of the vertical.”). In order to allow the horse to bend on the circle, the lungers typically adjusted the inside rein to be slightly shorter than the outside rein. The difference in length between the two sides was at the discretion of the lunger. A duration of 1 minute 30 sec of left canter without vaulter was recorded. The horses worked on a circle with a 14 ± 1 m long lunge line. The mean speed of the left canter was 4± 0.5 m/s, and the mean stride frequency was 1.6 ± 0.1 stride/s. The head and neck positions showed a head-neck angle of 91˚± 9 and a neck-trunk angle of 178˚ ± 9. These values were measured from the zygomatic arch, the withers, and base of the tail taken from still images from the camera worn by the lunger, for a canter stride. As described in the FEI rules, the nose was near the front of the vertical axis, at -2˚ ± 9. Ethical statement The study was approved by the FEI (Vaulting) and did not require study-specific permission. The behavioural observations of the horses were carried out in compliance with the ethical guidelines of the International Society for Applied Ethology. This research was non-invasive, and the test setup corresponded to a regular training session. The reins tested were part of the horses’ normal equipment. The tension was measured with a sensor fixed at the extremity of the reins. Chief trainers of the testing stations and the owners, the lungers and the vaulters volunteered their horses to participate in the study. They all signed an informed consent form beforehand. If the participant was a minor, one of the two parents signed. No personal data was collected. The age and number of years of practice were collected from the FEI website https://www.fei.org/vaulting/athletes). The data was anonymized for analysis. Mouth examination: An examination of the horse’s mouth before and after each test was carried out by a qualified FEI veterinarian equipped with disposable gloves. The examination consisted first of all of palpating the upper and lower bars, looking for sensitivity or injuries, and secondly, with the mouth open and the tongue held in place, checking for injuries to the mandibular bars, the lips, the corners and the tongue. These examinations ensured the respect for the physical integrity of the horse and the quality of the data collection. A form was filled in for each test, and the results were communicated live. No oral injuries were reported during the study. Data collection The judges’ scores A total of 66 videos (33 horses/2 types of side reins) were obtained. Five international judges scored the videos in a randomised order without knowing the name of the horse, the vaulter nor the nationality. Complete blinding was not possible. For each video, the judges gave 7 scores (Scores can range between 0-10 with 10 indicating the highest level, one decimal allowed). The parameters were scored as defined by the FEI Vaulting Guidelines (2021)): ● Rhythm: The characteristic sequence of footfalls and timing of a pure canter with clear beat and clear moment of suspension ● Relaxation: The Horse’s mental state (calmness without anxiety or nervousness), as well as its physical state (the absence of negative muscle tension). ● Connection: Through the body to the contact on the side reins (acceptance of the bit (or cavesson) through acceptance of the aids) when the energy generated in the hindquarters by the driving aids flows through the whole body of the Horse and is received in the side reins and lunge line, referred to as the ‘bridge of engagement’. ● Impulsion: The transmission of an eager and energetic, yet controlled, propulsive thrust generated from the hindquarters into the athletic movement of the Horse. Impulsion is associated with a phase of suspension. It is measured by the Horse’s engagement of its hind legs and desire to carry itself forward with a lowering of the croup, elasticity of its steps, suppleness of its back, and increased lightness of the forehand. ● Straightness: The Horse is straight, when the footfalls of the forehand and the hindquarters are appropriately aligned, and when its longitudinal axis is in line with the track on which it is lunged. ● Collection: The Horse lowers and engages its hindquarters, and shortens and narrows its base of support, resulting in lightness and mobility of the forehand and self-carriage. ● A1: the average of the six subscores listed above make up Score A1. A1 represents 60% of the total score for the horse. ● Vault-ability (A2): The Horse must give the impression of carrying out the Test of its own accord and stay in true balance and self-carriage. There is a picture of harmony and lightness throughout the Test. A2 represents 25% of the total score for the horse. Locomotor parameters The IMU fixed against the sternum was the preferred position for estimating movements of the centre of gravity that affect the distribution and intensity of loads on the (H. M. Clayton et Sha 2006) (Biau et Barrey 2004) The sensor records acceleration data on the z-axis (dorso-ventral direction of movement pointing upwards), the x-axis (horizontal direction of movement pointing forward) and the y-axis (lateral direction of movement pointing to the right), rotation angles around the three-dimensional axes (roll, pitch and yaw angles). For each of the 66 tests, the measurements described in table (1) were calculated from the IMU recordings. Fig 1: Representation of the three planes: Frontal plane in orange. This plane divides the body into front and back parts. In this plane, the horse performs movements of inclination and the oscillation around the longitudinal axis of the body is rolling. Sagittal plane in blue. This plane divides the body into right and left parts. In this plane, the horse performs movements of flexion (movement of the forehand downward) and extension (movement of the forehand upward) and the oscillation around the mediolateral axis is pitching. Transversal or horizontal plane in green. This plane divides the body into upper and lower parts. In this plane, the horse performs movements to the left and right and the oscillation around the vertical axis of the body is yawing. Statistical analyses The statistical analysis was carried out on 65 recordings including 2 recordings per horse (one with each type of side rein) and the seven scores from each judge for each recording. One horse was recorded with only one type of side reins. Descriptive statistics were carried out for the variables and Spearman’s correlation analyses were used to assess linear relationships between each variable, IMU sensor measurements, and judges’ scores. A classification using the Kmeans method on the canter (A1) + Vaultability (A2) score was carried out. A Mann-witney test (<0.05) was performed to determine if there was a significant difference in locomotor parameters between the two groups, one well rated by the judges’ ’good’ vs ’bad’ groups. All statistical analyses were performed using Xlstat 2023.3.0 (Lumivero (2025), XLSTAT statistical and data analysis solution. https://www.xlstat.com/fr) with a significance level of (p) < 0.05.

The distribution of the judges’ scores is shown in Fig 2. Only Collection and Connection were normally distributed (Shapiro test). A few scores were far from the median. Differences between the judges’ scores were not investigated. Fig 2: box plot of the judges’ scores. Descriptive statistics for locomotor parameters are presented table 2.

of the correlation tests between the variables are shown below, Fig 3. Fig 3 : Correlation plot. It displays significative correlations between the judges’ scores and the locomotor parameters. Positive correlations are in green and negative correlations are in red. Black means no correlation The correlation analysis showed that all the scores are highly correlated. All the scores were correlated with the same locomotor parameters and the quality index was inversely correlated with all the scores. Correlations for each of the scores with the locomotor parameters are described in Table 3. Correlation analysis showed a link between LgB and Pitch. A horse in extension was more stable (-0.247, 0.047), with more propulsion (-0.246, 0.049) and the more the horse propels, the higher the elevation (0.284,0.022). Obviously, propulsion and braking are linked: the more time the horse spends braking, the less time it spends pushing. Analysis of variance An analysis of variance (Mann and Witney) on the two groups formed from a k-means test on A1+A2 scores, differentiated good horses (‘Correct canter’ group) and less good horses (‘Incorrect canter’ group) by the elevation ( 14 ± 1 cm vs 13 ± 2cm, 0.013), the propulsion (6.7 ± 1.4 m/s² vs 5.7 ± 1.7 m/s², 0.005) and a clear regularity of the three beats (0.016 ± 0.004 m/s 3 vs 0.019 ± 0.004 m/s 3, 0.018) Fig 4, Fig 5. Fig 4: Boxplot of the three locomotor parameters that distinguished a correct canter quality from an incorrect one. Fig 5: FFT of a good canter quality index on the left and a bad one on the right. The yellow area represents the calculation of the canter quality index. The larger the area, the more irregular the canter.

Many sports, such as gymnastics, diving, ski jumping and figure skating, use judges’ scores to determine the winner of a competition. These judges use discipline specific rating scales when judging performances (for example, 0.000 - 10.000). This is the case for vaulting judges, and one judge (out of four judges) is responsible for scoring the Rhythm, Relaxation, Connection, Impulsion, Straighness and Collection of the canter with scores between 0-10, one decimal is allowed. The average of the six scores counts as A1. This A1 score is combined with the vaultability (A2) of the horse, which represents 25% of the horse’s final score, and the Lunging score, which represents 15% for a total Horse Score. A3 was not scored by the 5 judges in this study as the aim only concerned judging horse locomotion. The results showed a strong correlation between the 6 scores in A1 (+ A2): a horse judged to be good only received good marks. This may suggest that the judges assessed the overall quality of the canter rather than focusing on each of the 6 specific scores, or that the 6 scores are linked and thus if one goes up, the other scores follow. This refers to the complexity of judging without numerical data. The high number of canter quality criteria to be scored in 1 min by a single judge may give rise to bias (Dosseville et Garncarzyk 2007) (Plessner 1999) (Ste-Marie 2003) and requires a great deal of knowledge of horse locomotion. An assessment of seven (or more) scores without any numerical data has been reported to be beyond human cognitive capacity (Diaz et al. 2010). As in dressage, judges rely on simpler cognitive strategies to make decisions, which can lead to errors of judgement (Wolframm 2023) such as breed or the fact that some judges systematically give high marks and others low marks. According to the judges, a correct vaulting horse canter was distinguished from an incorrect one by the activity and the regularity of the three beats. Moreover, each score was correlated with the elevation and regularity of the three beats. Elevation and propulsion were two canter performance criteria that have already been demonstrated in dressage horses (9). Balance in sagittal and mediolateral planes can also be relevant observables. They correlated well with the judges’ scores. The horse has to show collection by lowering and engaging its hindquarters. The forehand is lightened as a result of the better balance; therefore, the horse will be able to move its forelimbs more freely, and reach higher and more forward. The collection of a gait means that the forward movements become upward movements. Propulsion and elevation were the consequence of a good balance. The horses with a lower leaning angle were scored better. Like dressage (H. M. Clayton et Sha 2006), trained vaulting horses learn to maintain the vertical orientation of their body in order to support the vaulter. This observable was correlated with the straightness score. Good straightness offered good stability in the mediolateral plane for the execution of the vaunter’s exercises. This correlation between all the scores may reflect the scale of progression proposed in the rules (1), which should make it possible to assess the quality of a performance objectively. The progression is organised into three main phases, the order of which is fundamental. Thus, a horse that is collected, the highest criterion on the scale, is balanced, with impulsion, and regular in its beats. According to this scale, a horse could have a poor score for collection, but a good score for regularity of the beats. However, this was not the case in this study: a horse with a poor collection score had systematically a poor regularity score. In addition, the guidelines (1) specify that scores must be distinct. For example, a horse may show poor rhythm with not a clear 3-beat canter (lower A1), but may still work with completely consistent speed and energy (higher A2). It seems unlikely that all 33 horses would have been either good or bad for all 6 scores, indicating that either the details of the 6 A1 scores is redundant, or the judges are not able to separate the judging of the scores. The fact that several of the scores were correlated with the same locomotor parameters showed that the assessment of these separately may be redundant and that a simplification such as reducing the number of scores might be possible. The quality index calculated in this study seemed to be of major significance to the judges. This index should reflect the distortions of the dorsoventral accelerometric signal and consequently evaluate a certain regularity of gait and even modifications of the three beats. During the test with the vaulter, the two major faults in the quality of the canter are a lack of projection and a desynchronization of the diagonal, which must be penalised by the ‘Rhythm’ score, which inevitably has an impact on the A1 score. This was the case in the study. As the other locomotor parameters were not correlated with this index, an in-depth biomechanical study of these two faults should be envisaged in order to detect them visually.

The main limitation in the current study design is that the scoring was done using videos. We know that the visual angle from which the sporting situation is perceived influences the judgement. Even if the camera were placed in the same place as a judge’s table, the information perceived visually is limited, particularly when it comes to assessing the speed of movements (Dosseville et Garncarzyk 2007). In addition, the attempted public pressure, which is part of the judging process, is not taken into account here. Scores can also be biased by the automation of certain evaluation routines, especially for a large number of videos, with an effect of the order in which the performances are viewed (Plessner 1999). As in the case of live judging, the participating judges were international level expert judges, and we know from previous studies in other disciplines that there is a risk of biases of expert judges, such as over-personalised expectations of some athletes highlighted in e.g. gymnastics, (Ste-Marie 2003). The videos given were not nominative, but the fact that they were international judges and international athletes made anonymising impossible. Measurements of canter quality were carried out with the vaulter, as in competition in this study. To extend this analysis, it would be relevant to compare the locomotor parameters of the canter with and without the vaulter. As with the rider-horse interaction, the vaulter probably has an impact on locomotion, whether beneficial or not (Peham et al. 2004) (MacKechnie-Guire et al. 2020).

The FEI rules propose 6 distinct scores to assess the quality of the canter. Rhythm, relaxation, connection, impulsion, straightness, and collection, and a separate score for vaultability. They were all correlated and therefore non-discriminatory, and several were correlated with the same locomotor parameters. On the basis of the results, a simplification of the number of scores can be discussed. The results showed that for judges, a good quality of canter was a canter with propulsion, elevation and a good regularity of the three beats. Balance in the sagittal and mediolateral planes should be retained as an observable of performance. All of these observables seem to be decisive and sufficient to assess the quality of the canter and are interesting from the point of view of decision support within the framework of the initial and ongoing training of judges. Références Biau, S., et Eric Barrey. 2004. « Relationship between stride characteristics and scores in dressage tests ». Pferdeheilkunde 20: 140‑44. doi:10.21836/PEM20040205. Clayton, H. M., et D. H. Sha. 2006. « Head and Body Centre of Mass Movement in Horses Trotting on a Circular Path ». Equine Veterinary Journal. Supplement (36): 462‑67. doi:10.1111/j.2042-3306.2006.tb05588.x. Clayton, Hilary M., et Sarah Jane Hobbs. 2017. « An Exploration of Strategies Used by Dressage Horses to Control Moments around the Center of Mass When Performing Passage ». PeerJ 5: e3866. doi:10.7717/peerj.3866. Diaz, Ana, Mary Johnston, Jennifer Lucitti, Wendi Neckameyer, et Katy Moran. 2010. « Scoring Variables and Judge Bias in United States Dressage Competitions ». Journal of Quantitative Analysis in Sports 6: 13‑13. doi:10.2202/1559-0410.1251. Dosseville, Fabrice, et Catherine Garncarzyk. 2007. « The Decision Making of Sports Referees ». Bulletin de psychologie 489(3): 225‑37. « FEI Vaulting Rules ». 2012. FEI . https://inside.fei.org/fei/disc/vaulting/rules (february 24th 2025). Lewczuk, D. 2013. « Effect of the Judge and Definition of the Trait for Horse Free Jumping Evaluation ». Archives Animal Breeding 56(1): 638‑49. doi:10.7482/0003-9438-56-064. MacKechnie-Guire, Russell, Erik MacKechnie-Guire, Vanessa Fairfax, Mark Fisher, Sophie Hargreaves, et Thilo Pfau. 2020. « The Effect That Induced Rider Asymmetry Has on Equine Locomotion and the Range of Motion of the Thoracolumbar Spine When Ridden in Rising Trot ». Journal of Equine Veterinary Science 88: 102946. doi:10.1016/j.jevs.2020.102946. Peham, C., T. Licka, H. Schobesberger, et E. Meschan. 2004. « Influence of the Rider on the Variability of the Equine Gait ». Human Movement Science 23(5): 663‑71. doi:10.1016/j.humov.2004.10.006. Plessner, Henning. 1999. « Expectation Biases in Gymnastics Judging ». Journal of Sport and Exercise Psychology 21(2): 131‑44. doi:10.1123/jsep.21.2.131. Ste-Marie, Diane. 2003. « Memory biases in gymnastic judging: Differential effects of surface feature changes ». Applied Cognitive Psychology 17: 733‑51. doi:10.1002/acp.897. Wolframm, Inga. 2023. « Let Them Be the Judge of That: Bias Cascade in Elite Dressage Judging ». Animals 13(17): 2797. doi:10.3390/ani13172797. Table 1: Canter calculated variables | Mean longitudinal balance: LgB (°) | Mean of body angle in the sagittal plane (in relation to the position before the start). It represents the average balance between the shoulders and the hips of the horse per stride; with positive values representing a flexion of the body (“balance on the shoulders”), and negative values representing an extension of the body (“balance on the hips”). (Fig.1) | Lightness of the forehand, with the poll as the highest point. | | Mean lateral balance: LaB (°) | Mean of body angle in relation to the position before the start. It represents the average left/right balance of the horse per stride; with negative values representing a lean to the left and positive values representing a lean to the right | Position and movement in the horizontal plane most closely correspond to the straightness score. The guidelines give an observation of the mean lateral balance at an almost vertical body angle. The results should therefore show a link between straightness and mean lateral balance. | | Roll body lean angle: Roll (°) | Roll angle of the rotation around the longitudinal axis. | | | Pitch body angle: Pitch (°) | Pitch angle of the rotation around the mediolateral axis. | Guidelines describes a stability in balance therefore a low pitch value | | Elevation (cm) | Range of vertical displacement of the horse’s body per stride (by the method for deriving displacement, Pfau, 2005) | Clear moment of suspension. Suppleness through the back. | | Propulsion (m/s²) | Mean vector of propulsion during stance phase (when dorsoventral acceleration was negative and longitudinal acceleration positive) | Guidelines describe a canter with energetic strides, clear impulsion from the engaged hindquarters. | | Braking (m/s²) | Mean vector of braking during stance phase (when both accelerations were negative) | | | Braking time (s) | Mean time spent braking during the support phase per stride | | | Propulsion time (s) | Mean time spent in propulsion during the support phase per stride. | | | Canter quality index (m/s 3 ) | Integral of a discrete Fourier transform of the first two harmonics ± frequency/2) (Fig 4). The lower the value, the more regular the three beats. A high value could represent a diagonal dissociation, i.e. a short temporal separation of the limbs of the diagonal pairs at contact and/or lift off. | Regularity and clear 3-beat. | Table 2: Range (minimum and maximum) and mean (max_min) for locomotor parameters | LgB (°) | -8 (-24 — 6) | | | LaB (°) | -11 (-24—0) | | | Pitch (°) | 13 (9 –19) | | | Roll (°) | 10 (6—17) | | | Braking time (s) | 0.18 (0.1—0.25) | | | Propulsion time (s) | 0.14 (0.05—0.24) | | | Braking (m/s²) | 7 (5.3—9.5) | | | Propulsion (m/s²) | 6.5 (3.5—13.2) | | | Elevation (cm) | 14 (9—17) | | | Quality index (m/s 3 ) | 0.017 (0.009—0.027) | Table 3: Correlations between scores and locomotor parameters. Only significant correlations (p<0.05) are reported. | Index quality | -0.550 (<0.0001) | -0.502 (<0.0001) | -0.441 (<0.0001) | -0.513 (<0.0001) | -0.467 (<0.0001) | -0.441 (<0.0001) | -0.504 (<0.0001) | -0.564 (<0.0001) | | LgB | -0.269 (0.030) | -0.258 (0.038) | -0.266 (0.033) | -0.252 (0.043) | |||| | Lab | 0.245 (0.050) | 0.303 (0.015) | 0.372 (0.002) | 0.384 (0.013) | 0.324 (0.009) | 0.330 (0.008) | 0.360 (0.003) | | | Elevation | 0.529 (<0.0001) | 0.301 (0.015) | 0.377 (0.002) | 0.525 (<0.0001 | 0.276 (0.027) | 0.456 (<0.0001) | 0.411 (0.001) | 0.369 (0.003) | | Propulsion | 0.305 (0.014) | 0.322 (0.009) | 0.269 (0.030) | 0.251 (0.044) | 0.279 (0.025) | ||| | Roll | -0.282 (0.002) | Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Metrics & Citations Metrics Article Usage 213views 115downloads Citations Download citation Sophie Biau, Elena Pycik, Robert de Bruin, et al. Alignment between gait parameters and judging of canter quality in international vaulting horses. Authorea. 22 May 2025. DOI: https://doi.org/10.22541/au.174791037.75207684/v1 DOI: https://doi.org/10.22541/au.174791037.75207684/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.