Classifying Winning Performances in International Women's Rugby Union.

PURPOSE: The efficacy of isolated and relative performance indicators (PIs) has been compared in rugby union; the latter more effective at discerning match outcomes. However, this methodology has not been applied in women's rugby. The aim of this study was to identify PIs that maximize prediction accuracy of match outcome, from isolated and relative data sets, in women's rugby union. METHODS: Twenty-six PIs were selected from 110 women's international rugby matches between 2017 and 2022 to form an isolated data set, with relative data sets determined by subtracting corresponding opposition PIs. Random forest classification was completed on both data sets, and feature selection and importance were used to simplify models and interpret key PIs. Models were used in prediction on the 2021 World Cup to evaluate performance on unseen data. RESULTS: The isolated full model correctly classified 75% of outcomes (CI, 65%-82%), whereas the relative full model correctly classified 78% (CI, 69%-86%). Reduced respective models correctly classified 74% (CI, 65%-82%) and 76% (CI, 67%-84%). Reduced models correctly predicted 100% and 96% of outcomes for isolated and relative test data sets, respectively. No significant difference in accuracy was found between data sets. In the relative reduced model, meters made, clean breaks, missed tackles, lineouts lost, carries, and kicks from hand were significant. CONCLUSIONS: Increased relative meters made, clean breaks, carries, and kicks from hand and decreased relative missed tackles and lineouts lost were associated with success. This information can be utilized to inform physical and tactical preparation and direct physiological studies in women's rugby.

Enhancing the Initial Acceleration Performance of Elite Rugby Backs. Part II: Insights From Multiple Longitudinal Individual-Specific Case-Study Interventions.

PURPOSE: This study implemented 18-week individual-specific sprint acceleration training interventions in elite male rugby backs based on their predetermined individual technical needs and evaluated the effectiveness of these interventions. METHODS: Individual-specific interventions were prescribed to 5 elite rugby backs over an 18-week period. Interventions were informed by the relationships between individual technique strategies and initial acceleration performance, and their strength-based capabilities. Individual-specific changes in technique and initial acceleration performance were measured at multiple time points across the intervention period and compared with 3 control participants who underwent their normal training. RESULTS: Of the technique variables intentionally targeted during the intervention period, moderate to very large (|d| = 0.93-3.99) meaningful changes were observed in the participants who received an individual-specific intervention but not in control participants. Resultant changes to the intervention participants' whole-body kinematic strategies were broadly consistent with the intended changes. Moderate to very large (|d| = 1.11-2.82) improvements in initial acceleration performance were observed in participants receiving individual-specific technical interventions but not in the control participants or the participant who received an individual-specific strength intervention. CONCLUSIONS: Individual-specific technical interventions were more effective in manipulating aspects of acceleration technique and performance compared with the traditional "one-size-fits-all" approach adopted by the control participants. This study provides a novel, evidence-based approach for applied practitioners working to individualize sprint-based practices to enhance acceleration performance.

Enhancing the Initial Acceleration Performance of Elite Rugby Backs. Part I: Determining Individual Technical Needs.

PURPOSE: This study sought to quantify the within-individual relationships between spatiotemporal variables and initial acceleration sprint performance in elite rugby backs and to establish a normative data set of relevant strength-based measures. METHODS: First, the spatiotemporal variables, ratios of step length to step rate and of contact time to flight time, and initial acceleration performance were obtained from 35 elite male rugby backs (mean [SD] age 25 [3] y) over the first 4 steps of 3 sprints. Angular and linear kinematic aspects of technique and strength-based qualities were collected from 25 of these participants. Second, the same spatiotemporal variables were collected from 19 of the participants on 3 further occasions (12 trials in total) to determine the within-individual associations of these variables and initial acceleration performance. RESULTS: Moderate to very large meaningful within-individual relationships (|r| = .43-.88) were found between spatiotemporal variables and initial acceleration performance in 17 of the 19 participants. From these relationships, a theoretically "desirable" change in whole-body kinematic strategy was individually determined for each participant, and normative strength-based measures to contextualize these were established. CONCLUSIONS: Meaningful within-individual relationships are evident between sprint spatiotemporal variables and initial acceleration performance in elite rugby backs. Individualized approaches are therefore necessary to understand how aspects of technique relate to initial acceleration performance. This study provides an objective, evidence-based approach for applied practitioners to identify the initial acceleration technical needs of individual rugby backs.

Lower-limb wearable resistance overloads joint angular velocity during early acceleration sprint running.

Lower-limb wearable resistance (WR) facilitates targeted resistance-based training during sports-specific movement tasks. The purpose of this study was to determine the effect of two different WR placements (thigh and shank) on joint kinematics during the acceleration phase of sprint running. Eighteen participants completed maximal effort sprints while unloaded and with 2% body mass thigh- or shank-placed WR. The main findings were as follows: 1) the increase to 10 m sprint time was small with thigh WR (effect size [ES] = 0.24), and with shank WR, the increase was also small but significant (ES = 0.33); 2) significant differences in peak joint angles between the unloaded and WR conditions were small (ES = 0.23-0.38), limited to the hip and knee joints, and <2° on average; 3) aside from peak hip flexion angles, no clear trends were observed in individual difference scores; and, 4) thigh and shank WR produced similar reductions in average hip flexion and extension angular velocities. The significant overload to hip flexion and extension velocity with both thigh- and shank-placed WR may be beneficial to target the flexion and extension actions associated with fast sprint running.

Understanding the effects of ball orientation in Rugby Union place kicking: the preferences of international kickers and the kinematics of the foot-ball impact.

Rugby Union place kicking is influential to match outcome. Previous research has analysed kicker motion prior to ball contact in detail, but ball orientation and the impact phase are typically ignored. This study aims to firstly identify the ball orientations used by international place kickers, and secondly to experimentally analyse the foot-ball interaction in trained kickers using different ball orientations. Overall, 25.5% of the international kickers used an upright ball orientation, 27.5% used a diagonal orientation and 47.1% used a horizontal orientation. However, ball orientation preference was not significant in predicting kick outcome in a binomial logistic regression model. To address the second aim, ball orientation was experimentally manipulated and lower limb and ball kinematics were captured using high-speed (4000 Hz) video. Whilst the impact location on the ball differed significantly between most ball orientation conditions, the impact location relative to the global vertical was largely consistent across all conditions. This was likely due to kickers adopting very consistent lower limb kinematics, although the shank and ankle angles at impact were affected by ball orientation conditions for some kickers. Impact durations also differed between some conditions, although this did not appear to affect the impact efficiency.

Inter- and intra-limb coordination during initial sprint acceleration.

In complex movements, centre of mass translation is achieved through effective joint and segment rotations. Understanding segment organisation and coordination is therefore paramount to understanding technique. This study sought to comprehensively describe inter- and intra-limb coordination and assess step-to-step changes and between-individual variation in coordination during initial sprint acceleration. Twenty-one highly trained to world class male (100 m PB 9.89-11.15 s) and female (100 m PB:11.46-12.14 s) sprinters completed sprint trials of at least 20 m from which sagittal plane kinematics were obtained for the first four steps using inertial measurement units (200 Hz). Thigh-thigh, trunk-shank and shank-foot coordination was assessed using a modified vector coding and segment dominancy approach. Common coordination patterns emerged for all segment couplings across sexes and performance levels, suggesting strong task constraints. Between-individual variation in inter-limb thigh coordination was highest in early flight, while trunk-shank and shank-foot variation was highest in late flight, with a second peak in late stance for the trunk-shank coupling. There were clear step-to-step changes in coordination, with step 1 being distinctly different to subsequent steps. The results demonstrate that inter-limb coordination is primarily anti-phase and trailing leg dominant while ankle motion in flight and late stance appears to be primarily driven by the foot.

Determining jumping performance from a single body-worn accelerometer using machine learning.

External peak power in the countermovement jump is frequently used to monitor athlete training. The gold standard method uses force platforms, but they are unsuitable for field-based testing. However, alternatives based on jump flight time or Newtonian methods applied to inertial sensor data have not been sufficiently accurate for athlete monitoring. Instead, we developed a machine learning model based on characteristic features (functional principal components) extracted from a single body-worn accelerometer. Data were collected from 69 male and female athletes at recreational, club or national levels, who performed 696 jumps in total. We considered vertical countermovement jumps (with and without arm swing), sensor anatomical locations, machine learning models and whether to use resultant or triaxial signals. Using a novel surrogate model optimisation procedure, we obtained the lowest errors with a support vector machine when using the resultant signal from a lower back sensor in jumps without arm swing. This model had a peak power RMSE of 2.3 W·kg-1 (5.1% of the mean), estimated using nested cross validation and supported by an independent holdout test (2.0 W·kg-1). This error is lower than in previous studies, although it is not yet sufficiently accurate for a field-based method. Our results demonstrate that functional data representations work well in machine learning by reducing model complexity in applications where signals are aligned in time. Our optimisation procedure also was shown to be robust can be used in wider applications with low-cost, noisy objective functions.

Acute effects of wearable thigh and shank loading on spatiotemporal and kinematic variables during maximum velocity sprinting.

Light wearable resistance is used in sprint training but the scientific evidence to guide its implementation is limited. This study investigated thigh and shank loading protocols which were matched based on the average increase in moment of inertia about the hip over a stride cycle. Seven university-level sprinters completed three counterbalanced conditions (unloaded, shank-loaded, thigh-loaded), and kinematic variables were measured between 30 and 40 m. Both thigh and shank loading led to small reductions in step velocity (mean change = -1.4% and -1.2%, respectively). This was due to small reductions in step frequency (-1.8%; -1.7%) because of small increases in contact time (+2.7%; +1.5%) in both conditions and a small increase in flight time (+2.0%) in the shank-loaded condition. Both conditions led to moderate increases in hip extension at toe-off (+2.7°; +1.4°), whilst thigh loading led to a small reduction in peak hip flexion angle during swing (-2.5°) and shank loading led to a small increase in peak biceps femoris muscle-tendon unit length (+0.4%). Thigh and shank loading can both be used to provide small reductions in sprint velocity, and each has specific overload effects which must be considered in the rationale for their implementation.

Characterising initial sprint acceleration strategies using a whole-body kinematics approach.

Sprint acceleration is an important motor skill in team sports, thus consideration of techniques adopted during the initial steps of acceleration is of interest. Different technique strategies can be adopted due to multiple interacting components, but the reasons for, and performance implications of, these differences are unclear. 29 professional rugby union backs completed three maximal 30 m sprints, from which spatiotemporal variables and linear and angular kinematics during the first four steps were obtained. Leg strength qualities were also obtained from a series of strength tests for 25 participants, and 13 participants completed the sprint protocol on four separate occasions to assess the reliability of the observed technique strategies. Using hierarchical agglomerative cluster analysis, four clear participant groups were identified according to their normalised spatiotemporal variables. Whilst significant differences in several lower limb sprint kinematic and strength qualities existed between groups, there were no significant between-group differences in acceleration performance, suggesting inter-athlete technique degeneracy in the context of performance. As the intra-individual whole-body kinematic strategies were stable (mean CV = 1.9% to 6.7%), the novel approach developed and applied in this study provides an effective solution for monitoring changes in acceleration technique strategies in response to technical or physical interventions.

The approach towards the ball, rather than the physical characteristics of the kicker, limits accurate rugby place kicking range.

The aim of this study was to understand how a place kicker's range is limited by their approach to the ball and their physical characteristics. Thirty-three kickers performed maximal place kicks and vertical jumps in a laboratory. Whole-body motion and ground reaction forces during the approach phase of the kicks, jump performance and anthropometric measurements of those whose predicted maximum distance was limited by range (n = 17) rather than accuracy were analysed. Principal component analysis (PCA) reduced the number of variables considered before stepwise regression analyses assessed variance in place kick maximum distance and associated criteria. Four components, explaining 94% of the variance in maximum distance, were extracted from the PCA: width of approach, anterior-posterior body position, centre-of-mass height and lower limb strength. Lower limb strength was a significant predictor of both kicking foot velocity (R2 = 0.55, p = 0.001) and ball velocity magnitude (R2 = 0.57, p < 0.001). However, maximum distance was determined by body position during the approach (antero-posterior position, R2 = 0.52, p = 0.001 and centre-of-mass height, R2 = 0.12, p = 0.049). This highlights the importance of considering three-dimensional motion of the kicker alongside their physical capabilities to understand place kicking range.

Relationships between kinematic characteristics and ratio of forces during initial sprint acceleration.

In track sprinting, acceleration performance is largely determined by the ability to generate a high ratio of forces (RF), but the technical features associated with this remain unknown. This study therefore investigated the relationships between selected kinematic characteristics and RF during the initial acceleration phase. Fourteen male sprinters completed two maximal 60 m sprints from a block start. Full-body kinematic and external kinetic data were obtained from the first four steps, and the relationships between selected kinematic characteristics and mean RF over the first four steps were determined. Placing the stance foot further behind (or less far in front of) the whole-body centre of mass at touchdown was significantly related to greater RF (r = -0.672), and more anterior orientation of the proximal end of the foot (r = -0.724) and shank (r = -0.764) segments at touchdown were also significantly related to greater RF. Following touchdown, greater ankle dorsiflexion range of motion during early stance was significantly related to greater RF (r = 0.728). When aiming to enhance RF during initial acceleration, practitioners should be encouraged to focus on lower leg configurations when manipulating touchdown distance, and the role of dorsiflexion during early stance is also an important consideration.

Effects of manipulating specific individual constraints on performance outcomes, emotions, and movement phase durations in Rugby Union place kicking.

Place kicks present valuable opportunities to score points in Rugby Union, contributing almost half of all points scored at international level. From an ecological dynamics perspective, place kickers adapt to interacting task, environmental, and individual constraints in performance environments. The aim of this study was to analyse effects of specific manipulations of individual constraints (fatigue; expectation for success) on place kicking performance, movement phase durations, heart rate and self-reported emotions. Under representative task and environmental constraints on an outdoor training pitch, 12 experienced, male place kickers completed four testing sessions under every combination of manipulated high/low expectation for success and manipulated levels of high/low acute fatigue. Within each session of 12 place kicks, performance outcomes from three kicking locations of varying difficulty were recorded. ANOVA revealed a two-way interaction between fatigue and expectation manipulations on mean success percentage (p < 0.05), with higher success under low fatigue + low expectation (M ± SD = 58 ± 13%) and high fatigue + high expectation (M ± SD = 56 ± 14%), compared with separate manipulations of high expectation (M ± SD = 49 ± 14%) or high fatigue (M ± SD = 51 ± 14%). There were no significant effects on any movement phase durations. Manipulating expectation significantly heightened mean heart rate (p < 0.05) and influenced emotions reported by place kickers, including higher anger scores when there was high expectation for success. Coaches are encouraged to integrate place kicking into representative game scenarios in practice environments to faithfully represent key performance constraints (e.g. fatigue; expectation for success) in preparing kickers for competitive situations.

Changes to horizontal force-velocity and impulse measures during sprint running acceleration with thigh and shank wearable resistance.

This study determined the effects of two wearable resistance (WR) placements (i.e. thigh and shank) on horizontal force-velocity and impulse measures during sprint running acceleration. Eleven male athletes performed 50 m sprints either unloaded or with WR of 2% body mass attached to the thigh or shank. In-ground force platforms were used to measure ground reaction forces and determine dependent variables of interest. The main findings were: 1) increases in sprint times and reductions in maximum velocity were trivial to small when using thigh WR (0.00-1.93%) and small to moderate with shank WR (1.56-3.33%); 2) athletes maintained or significantly increased horizontal force-velocity mechanical variables with WR (effect size = 0.32-1.23), except for theoretical maximal velocity with thigh WR, and peak power, theoretical maximal velocity and maximal ratio of force with shank WR; 3) greater increases to braking and vertical impulses were observed with shank WR (2.72-26.3% compared to unloaded) than with thigh WR (2.17-12.1% compared to unloaded) when considering the entire acceleration phase; and, 4) no clear trends were observed in many of the individual responses. These findings highlight the velocity-specific nature of this resistance training method and provide insight into what mechanical components are overloaded by lower-limb WR.

Calculation of the centre of pressure on the athletic starting block.

We aimed to evaluate the accuracy of a new method to calculate the centre of pressure (COP) on a starting block above a force platform, and to examine how this method affected lower extremity joint torques during the block clearance phase compared against a previously used method which projects the COP from the metatarsophalangeal (MP) joint. To evaluate the accuracy of the new method, one experimenter applied force at 18 known locations on a starting block (under six block position and orientation conditions), during which ground reaction force was recorded underneath using a force platform. Two sprinters then performed three block starts each, and lower extremity joint torques were calculated during block clearance using the COP obtained from the new method and from the projection of the MP joint location. The calculated COP using the new method had a mean bias of ≤0.002 m. There were some large differences (effect sizes = 0.11-4.01) in the lower extremity joint torques between the two methods which could have important implications for understanding block clearance phase kinetics. The new method for obtaining the COP on a starting block is highly accurate and affects the calculation of joint torques during the block clearance phase.

Predicting performance at the group-phase and knockout-phase of the 2015 Rugby World Cup.

The primary aim of this paper was to produce a model that predicts outcome in the group-phase of the 2015 Rugby World Cup and to determine the relevance and importance of performance indicators (PIs) that are significant in predicting outcome. A secondary aim investigated whether this model accurately predicted match outcome in the knockout-phase of the competition. Data was the PIs from the 40 group-phase games of the 2015 RWC. Given the binary outcome (win/lose), a random forest classification model was built using the data sets. The outcome of the knockout-phase was predicted using this model and accuracy of prediction of the model from the group-phase. The model indicated that thirteen PIs were significant in predicting match outcome in the group-phase and provided accurate prediction of match outcome in the knockout-phase. These PIs were tackle-ratio, clean breaks, average carry, lineouts won, penalties conceded, missed tackles, lineouts won in the opposition 22, defenders beaten, metres carried, kicks from hand, lineout success, penalties in opposition 22 m and scrums won. For the group-phase matches tackle ratio, clean breaks and average carry were accurate standalone predictors of match outcome and respectively predicted 75%, 70% and 73% of match outcomes. The model based on the group-phase predicted correctly 7 from 8 (87.5%) knockout-phase matches. In the knockout-phase clean breaks predicted 7 from 8 outcomes, whilst tackle ratio and average carry predicted 6 from 8 outcomes.

Wearable resistance sprint running is superior to training with no load for retaining performance in pre-season training for rugby athletes.

This study determined the effects of a six-week lower-limb wearable resistance training (WRT) intervention on sprint running time, velocity, and horizontal force-velocity mechanical variables. Twenty-two collegiate/semi-professional rugby athletes completed pre- and post-intervention testing of three maximal effort 30 m sprints. A radar device was used to measure sprint running velocity from which horizontal force-velocity mechanical profiling variables were calculated. All athletes completed two dedicated sprint training sessions a week for six-weeks during pre-season. The intervention (wearable resistance, WR) group completed the sessions with 1% body mass load attached to the left and right shanks (i.e. 0.50% body mass load on each limb), whilst the control group completed the same sessions unloaded. For the control group, all variables were found to detrain significantly (p ≤ 0.05) over the training period with large detraining effects (ES > 0.80) for theoretical maximal horizontal force, slope of the force-velocity profile, maximal ratio of force, index of force application, 5 and 10 m times. For the WR group, there were no significant changes to any recorded variables (all p > 0.05) and all effects of training were trivial or small (ES < 0.50). After adjustment for baseline differences, significant between group differences were found for all variables (large effects, ES > 0.80) except theoretical maximal velocity, 30 m time, and maximal velocity. The addition of light wearable resistance to sprint training during a six-week pre-season block enables the maintenance of sprint performance and mechanical output qualities that otherwise would detrain due to inadequate training frequencies.

The importance of duration and magnitude of force application to sprint performance during the initial acceleration, transition and maximal velocity phases.

Successful sprinting depends on covering a specific distance in the shortest time possible. Although external forces are key to sprinting, less consideration is given to the duration of force application, which influences the impulse generated. This study explored relationships between sprint performance measures and external kinetic and kinematic performance indicators. Data were collected from the initial acceleration, transition and maximal velocity phases of a sprint. Relationships were analysed between sprint performance measures and kinetic and kinematic variables. A commonality regression analysis was used to explore how independent variables contributed to multiple-regression models for the sprint phases. Propulsive forces play a key role in sprint performance during the initial acceleration (r = 0.95 ± 0.03) and transition phases (r = 0.74 ± 0.19), while braking duration plays an important role during the transition phase (r = -0.72 ± 0.20). Contact time, vertical force and peak propulsive forces represented key determinants (r = -0.64 ± 0.31, r = 0.57 ± 0.35 and r = 0.66 ± 0.30, respectively) of maximal velocity phase performance, with peak propulsive force providing the largest unique contribution to the regression model for step velocity. These results clarified the role of force and time variables on sprinting performance.

The epidemiology of kicking injuries in professional Rugby Union: A 15-season prospective study.

PURPOSE: While kicking in Rugby Union can be influential to match outcome, the epidemiology of kicking injuries remains unknown. This study therefore aimed to investigate the epidemiology of injuries attributed to kicking in professional rugby, including playing position-specific effects and differences in kicking volumes and kick types. METHODS: Fifteen seasons of injury surveillance data and two seasons of match kicking characteristics from professional rugby players were analyzed. Incidence, propensity, and severity of kicking-related injuries were calculated together with the locations and types of these injuries. Position-related differences in match kicking types and volumes were also established. RESULTS: Seventy-seven match and 55 training acute-onset kicking injuries were identified. The match kicking injury incidence for backs was 1.4/1000 player-match-hours. Across all playing positions, the propensity for match kicking injury was 0.57 injuries/1000 kicks. Fly-halves sustained the greatest proportion of match kicking injuries (47%) and performed the greatest proportion of match kicks (46%); an average propensity for match kicking injury (0.58/1000 kicks). Scrum-halves executed 27% of match-related kicks but had a very low propensity for match kicking injury (0.17/1000 kicks). All other positional groups executed a small proportion of match-related kicks but a high propensity for match kicking injury. Ninety-two percent of match kicking injuries occurred in the pelvis or lower limb, with the majority sustained by the kicking limb. 21% of all match kicking injuries were associated with the rectus femoris muscle. CONCLUSION: Match kicking profiles and kicking injuries sustained are position-dependent, which provides valuable insight for developing player-specific conditioning and rehabilitation protocols.

Phase analysis in maximal sprinting: an investigation of step-to-step technical changes between the initial acceleration, transition and maximal velocity phases.

The aim of this study was to investigate spatiotemporal and kinematic changes between the initial acceleration, transition and maximum velocity phases of a sprint. Sagittal plane kinematics from five experienced sprinters performing 50-m maximal sprints were collected using six HD-video cameras. Following manual digitising, spatiotemporal and kinematic variables at touchdown and toe-off were calculated. The start and end of the transition phase were identified using the step-to-step changes in centre of mass height and segment angles. Mean step-to-step changes of spatiotemporal and kinematic variables during each phase were calculated. Firstly, the study showed that if sufficient trials are available, step-to-step changes in shank and trunk angles might provide an appropriate measure to detect sprint phases in applied settings. However, given that changes in centre of mass height represent a more holistic measure, this was used to sub-divide the sprints into separate phases. Secondly, during the initial acceleration phase large step-to-step changes in touchdown kinematics were observed compared to the transition phase. At toe-off, step-to-step kinematic changes were consistent across the initial acceleration and transition phases before plateauing during the maximal velocity phase. These results provide coaches and practitioners with valuable insights into key differences between phases in maximal sprinting.

A joint kinetic analysis of rugby place kicking technique to understand why kickers achieve different performance outcomes.

We aimed to identify differences in kicking leg and torso mechanics between groups of rugby place kickers who achieve different performance outcomes, and to understand why these features are associated with varying levels of success. Thirty-three experienced place kickers performed maximum effort place kicks, whilst three-dimensional kinematic (240 Hz) and ground reaction force (960 Hz) data were recorded. Kicking leg and torso mechanics were compared between the more successful ('long') kickers and two sub groups of less successful kickers ('short' and 'wide-left') using magnitude-based inferences and statistical parametric mapping. Short kickers achieved substantially slower ball velocities compared with the long kickers (20.8 ±â€¯2.2 m/s vs. 27.6 ±â€¯1.7 m/s, respectively) due to performing substantially less positive hip flexor (normalised mean values = 0.071 vs. 0.092) and knee extensor (0.004 vs. 0.009) joint work throughout the downswing, which may be associated with their more front-on body orientation, and potentially a lack of strength or intent. Wide-left kickers achieved comparable ball velocities (26.9 ±â€¯1.6 m/s) to the long kickers, but they were less accurate due to substantially more longitudinal ball spin and a misdirected linear ball velocity. Wide-left kickers created a tension arc across the torso and therefore greater positive hip flexor joint work (normalised mean = 0.112) throughout the downswing than the long kickers. Whilst this may have assisted kicking foot velocity, it also induced greater longitudinal torso rotation during the downswing, and may have affected the ability of the hip to control the direction of the foot trajectory.