The penalty corner is one of the most important game situations in field hockey, with one third of the goals resulting from this tactical situation (Laird and Sutherland, 2003; Piñeiro, 2008). The drag-flick is between 1.4 and 2.7 times more efficient than hitting or push-shooting the ball towards the goal when playing a penalty corner (McLaughlin, 1997; Piñeiro et al., 2007; Yusoff et al., 2008).

Only a few studies have analyzed the drag-flick. Some of them have provided kinematic information about players from different levels (McLaughlin, 1997; Yusoff et al., 2008; López de Subijana et al., 2010). These authors reported the cues which indicated a drag-flick: a wide stance, a whipping action of the stick before the hips and shoulders were rotated, and a final acceleration of the stick. In addition, Baker et al., 2009 focused on anticipation skills of the goalkeepers, while Jennings et al., 2010 studied the registered forces on the face of the stick. All of these studies were descriptive in nature.

Most of the previous field hockey experimental studies have focussed on training topics, such as endurance (Manna et al., 2009; Chapman et al., 2009), general physical condition (Astorino et al., 2004; Spencer et al., 2004), velocity (Bloomfield et al., 2007) and strength (Cochrane and Stannard, 2005). In relation to technical training, Beckamnn et al., 2010 applied different treatments for the push and the flick in indoor hockey twice per week during six weeks, obtaining very heterogeneous findings. To date, no studies have been conducted concerning the training of the drag-flick skill in field hockey.

Therefore, the aim of this study was to develop and apply a training method, based on previous studies, to improve the drag-flick skill on a young top-class field hockey player.

One male drag-flicker (19 years old; 66.8 kg; 1.71 m; eight years of field hockey experience) participated in this study. He was a drag- flicker from the under-21 Spanish National Team. The participant was requested to provide informed consent prior to his participation. The University's Ethics Committee approved the research protocol.

The training sessions were conducted in the hockey field of the Spanish Sports Council's High Performance Centre. The player exercised three times per week using specific drills over a four week period, completing a total of 12 sessions (Beckamnn et al., 2010). The average duration of the training sessions was 45 minutes and they were supervised by a qualified hockey coach and ex-Olympic athlete. The training sessions started with a preliminary warm up which was followed by four drills ordered by increasing complexity (Figures 1, 2, 3 and 4). Each drill was related to findings from previous studies (McLaughlin, 1997; Yusoff et al., 2008; López de Subijana et al., 2010) and it was performed in 2 sets with 7 repetitions per set (2x7). After each drill, 10 free drag- flicks were performed in order to add the new information to the overall movement. The training sessions were designed and organized according to a panel of expert. All the coaches selected had a minimum of 10 years experience as hockey coaches and they were members of the staff of the Royal Spanish Hockey Federation. Three dimensional (3D) data analysis was conducted prior to and after the training period.

All of the measurements were carried out in the Biomechanics Laboratory of the Faculty of Physical Activity and Sport Sciences at the Technical University of Madrid. A VICON optoelectronic system (Oxford Metrics, Oxford, UK) captured the drag-flicks with six cameras, sampling at 250 Hz. The experimental space was 5m long, 2.5m wide and 2m high, and was dynamically and statically calibrated with an error of less than 2 cm and a static reproducibility of 0.4%. A total of 49 retro- reflective markers (46 body markers and three 14 mm diameter stick markers) were attached to anatomical landmarks following VICON's kinematics model (Vicon Motion Systems, 2003). The stick markers were placed where the player's grip began, at the toe of the shaft, and at the end of the shaft. The stick's features (height: 94 cm; mass: 584.6 g; distance between the center of mass position and the end of the shaft: 38.4 cm) were approved by the International Hockey Federation (2009). Raw data were filtered using quintic spline functions based on Woltring's Generalized Cross-Validation (GCV) method for calculating the smoothing factor (Woltring, 1986). As markers could not be placed on the ball, an official field hockey ball was covered with adhesive reflective material. VICON cameras recognized the ball as a marker and ball velocity was estimated.

After a specific warm-up, 20 trials were carried out and captured at habitual speed. In each trial, the participant shot into a goal area marked with a fence in front of the player. If the participant did not score in the goal area, the trial was rejected. The ball was placed by the subject approximately 1.5 to 2 m away from the centre of the calibrated area. The drag-flick movement commenced once the front foot made contact with the floor, and finished 20 frames after the stick's peak positive angular velocity. The pelvis, upper trunk and stick angles were calculated using the line of the double foot contact as the y-axis, the x-axis as 90º to the right of the y-axis and the z-axis as the vertical axis. The angular velocities were computed from the angles formed by the upper trunk (shoulder line), pelvis (hip line), and stick with the x-axis on the xy plane. The knee flexion angle was computed for the front leg only. The knee flexion angle was computed for the front leg only.

The following key events of the drag-flick were identified: T1 (front foot contact); T2 (maximum angular velocity of the pelvis); T3 (peak negative angular velocity of the stick); T4 (maximum angular velocity of the upper trunk); T5 (maximum angular velocity of the stick); and T6 (ball release). The event times were normalized to the T1-T6 times. The stance width, drag-flick distance and the front foot-ball distance at T1 were normalized to the player's height.

Statistical analysis was carried out using SPSS v.16 software (SPSS Inc., Chicago, IL, United States). The means and standard deviations of the study parameters were calculated. In order to analyze pre- and post-test differences, a dependent t-test was carried out. The alpha level of significance was set at p < 0.05 for all statistical tests.

The player achieved ball velocities of 24.9 ± 0.9 m·s^-1 prior to the training period and 24.6 ± 0.8 m·s^-1 after the training period. The angular velocities are shown in Table 1. The angular velocities of the stick in the post-test improved (p > 0. 05) up to -256.0 ± 56.05 º/s the negative peak and 1315.4 ± 153.9 º/s the maximum peak. The maximum angular velocity of the upper trunk reached 475.3 ± 32.4 º/s after the training period (p > 0.05). Table 2 shows the kinematic sequence, which was similar before and after the training period. Peak angular velocity of the pelvis was recorded at 35.9 ± 5.4 and 38.9 ± 2.8% of the total time. The whipping effect of the stick was found close to 50% of the duration of the drag-flick (51.8 ± 4.4 and 50.8 ± 4.1 %). The maximum angular velocity of the stick was achieved near the release (101.3 ± 5.9 and 100.3 ± 5.1 %). Distance and angular parameters are shown in Table 3. The player increased 9 cm (p < 0.01) the distance from the front foot to the ball at T1 and 7 cm (p > 0.05) the drag-flick distance The range of motion from the front leg decreased from T1 to T6 after the training period (p < 0.01), from 37.7 ± 3.7º to 27.3 ± 3.9º, respectively.

The present study has shown improvements on the technical performance of a young top-class player, even though the training period was only four weeks long.The player achieved ball velocities which were close to those recorded by high performance players who have participated in previous studies (López de Subijana et al., 2010; Yusoff et al., 2008;), but lower than the 30.5 m·s^-1 measured using radar by Baker et al., 2009. The values reported for the negative peak of the angular velocity of the stick, the maximum angular velocity of the pelvis and the maximum angular velocity of the upper trunk were in the range suggested in a previous study (López de Subijana et al., 2010). The angular velocity of the stick was lower than that of a skilled drag-flicker (1890.1 ± 72.8 º/s) and the male´s group (1473.2 ± 177.8 º/s) from López de Subijana et al., 2010. This could be due to the age of the participant in this case study, as the player was young (19 years old), compared with the samples in previous studies (López de Subijana et al., 2010; Yusoff et al., 2008). The kinematic sequence was similar to that previously reported. Data concerning the stance width were similar to those found by McLaughin (1997), Yusoff et al., 2008 and López de Subijana et al., 2010. Drag-flick velocity was lower than the 11.6 m·s^-1 and 12.3 m·s^-1 recorded by López de Subijana et al., 2010 and McLaughin (1997), respectively.

The results of this study have shown improvements during the earlier phases of the skill..The player increased drag-flick front foot to the ball at double foot contact (T1) distances. These improvements could be due to drills number 1 (rolling the ball on the stick), 2 (taking the ball behind the rear foot), and 3 (making the last stance wider). The previous phase is very important in order to prepare a clear drag of the stick before the final acceleration. Complementarily, the player achieved a no statistically significant higher maximum angular velocity of the stick. The front leg flexion decreased after training, similarly to the skilled drag-flicker studied by López de Subijana et al., 2010.

A limitation of this study could be that there were two data collection, prior and after the training period. Beckamnn et al., 2010 also registered the precision six and twelve weeks later in order to measure the retention of the treatments.

• This article adds information about the drag-flick kinematics.
• This article adds information about how to train the drag-flick.
• The drag-flick is the most efficient technique shooting for goal after a penalty corner.