Spatiotemporal Analysis of Linear Skating Sprint in Male and Female Bandy Players: Analysis of Acceleration and Maximal Speed Phase

Abstract

Background/Objectives: This study aimed to investigate the skating determinants and differences between male and female bandy players in the spatiotemporal variables during acceleration and maximum sprint skating velocity. Methods: Seventy-four female bandy players (age: 18.9 ± 4.1 years; height: 1.67 ± 0.06 m; body mass: 63.2 ± 7.4 kg; training experience: 13.4 ± 3.9 yrs.; and 26 elite and 48 junior elite) and 111 male bandy players (age: 20.7 ± 5.0 years; height: 1.80 ± 0.05 m; body mass: 76.4 ± 8.4 kg; training experience: 13.8 ± 5.0 yrs.; and 47 elite and 66 junior elite players) performed linear sprint skating over 80 m. Split times were measured every ten metres by photocells to calculate velocities for each step and spatiotemporal skating variables (glide times and length, step length, and frequency) by IMUs attached to the skates. The first six steps (acceleration phase), the six steps at the highest velocity (maximal speed phase), and the average of all steps were used for analysing glide-by-glide spatiotemporal variables. Results: These revealed that male players exhibited higher acceleration and maximal skating velocity than female players. A higher acceleration in men was accompanied by shorter gliding time, longer step length, and higher step frequency. When skating at maximal speed, male players had a longer step length and gliding time and length. The sub-group analysis revealed that step frequency did not correlate with skating velocity, acceleration, or maximal speed phases. On the other hand, glide and step lengths significantly correlated with skating velocity in both phases (r ≥ 0.60). Conclusions: In general, for faster skating in bandy, it is generally better to prioritise glide and step length than stride frequency. Hence, players should be encouraged to stay low and have more knee flexion to enable a longer extension length and, therefore, a longer path and more horizontal direction of applied force to enhance their acceleration ability.

1. Introduction

Bandy is a team winter sport played in two halves of 45 min on a large, soccer-pitch-sized ice rink (i.e., 45–65 m × 90–110 m) between two teams of 11 players (including a goalkeeper) who skate and use their sticks, aiming to score by striking the ball into the opposing team’s goal [1]. It is a high-intensity intermittent sport that requires players to repeatedly engage in fast-paced skating activities (e.g., acceleration, sprinting, and changes of direction) that are interspersed by short bouts of active and passive recovery [1,2,3]. For instance, elite male players cover distances between 18 km and 25 km, spending between 40 min and 80 min of playing time at between 81% and 100% of their maximum heart rate and between 15 min and 27 min above their lactate threshold during a game [1,2]. As such, bandy requires both very well-developed aerobic and anaerobic capacities and lower-body power and strength [4].

Given that elite players skate ≈2.4 km very fast and ≈600 m while sprinting, sometimes reaching a speed over 37 km/h, it is reasonable to conclude that high-velocity skating and acceleration ability are two of the most important skills in bandy [1,2,3]. Recent studies have revealed that male and female elite senior bandy players could accelerate faster and reach a higher skating velocity than their junior counterparts [5,6]. Moreover, elite male players showed advanced skating speed and acceleration compared to elite female players [5,6], which partially could be attributed to well-documented sex differences in general (e.g., body height, weight, fat, lean mass percentage, strength, and power) [7,8,9]. However, it is insufficient to consider only physiological attributes to explain skating performance differences. Therefore, investigating skating techniques might be crucial to identify other factors that make the differences in skating speed and acceleration between the sexes [7,9].

Knowing that there is no research on the differences between male and female bandy players in terms of skating technique (i.e., kinematic parameters), we consider it appropriate to base our study’s hypotheses on existing research in ice hockey that shares similarities to bandy concerning the skating movement pattern [5,6]. Two recent studies showed clear differences between high-calibre male and female ice hockey players in skating technique and performance. In brief, both Shell, et al. [7] and Budarick, et al. [9] used a 3D motion capture system to investigate lower body mechanics during linear skating sprints, with a difference that Shell, et al. [7] analysed only the acceleration phase (i.e., the first seven steps), while Budarick, et al. [9] examined both acceleration and maximum speed phases over a 34 m distance. The authors showed indisputable sex differences, with males having a higher acceleration and achieving higher maximum skating speeds. This was followed by greater hip abduction and knee flexion at ice contact in the male than in the female players. Furthermore, the joint angles were greater during acceleration than the maximal skating phase in both cohorts, which warrants the investigation of both phases. This rationale is supported by other research that also has shown changes in spatiotemporal skating variables from a running-like technique at the acceleration to the gliding technique during the maximal speed [10,11,12]. For instance, a larger stride propulsion and push-off force were accompanied by shorter contact, stride and glide time, higher step frequency and higher muscle activation (e.g., the gastrocnemius) in the acceleration than in the maximal speed phase [10,11,12].

In addition to the above, there are some rule-based discrepancies between the sports that might affect acceleration and maximal speed. Therefore, the complete translation of findings from ice hockey research into bandy might be inappropriate and insufficient. In particular, players in bandy are less limited by the ice rink size (i.e., 90–110 m) to accelerate longer and reach a higher maximal speed compared to players in ice hockey, where the ice rink size is 60 m × 30 m, thereby, making bandy a faster sport expressed by higher maximal skating velocities. This means that Bandy places significantly greater demands on the ability to achieve a high maximum skating velocity compared to ice hockey and in comparison, ice hockey relies less on initial acceleration and more on rapid starts and stops. Furthermore, to support fast linear skating, Bandy skates have a significantly larger radius on the skate rail than ice hockey skates. Specifically, van den Tillaar and colleagues [5,6] reported a peak speed after 80 m of 9.52 ± 0.37 m/s and 10.83 ± 0.37 m/s in bandy female and male bandy players, respectively, which was higher than in ice hockey players (8.02 ± 0.36 m/s vs. 8.96 ± 0.44 m/s) tested over 34 m [9]. However, due to the marginal acceleration, the hockey players did not seem to reach their peak speed after 34 m [9]. It is known that rapid changes in direction and speed, and fast acceleration are crucial in ice hockey [13], while long-gliding periods and maximal speed are more important in bandy [5,6].

Although 3D motion capture systems are highly accurate and regarded as the ‘gold standard’ method for analysing spatiotemporal skating variables, they are often expensive, require specialised knowledge to operate, are not widely available in ice sports facilities, and have limited testing area coverage [14]. Specifically, in both studies, Shell, et al. [7] and Budarick, et al. [9] used an 18-camera motion capture system that was calibrated to capture an approximate skating area of 15 m long by 3 m wide and 2 m high. All of these factors reduce their practicality and usefulness. Therefore, we decided to investigate spatiotemporal kinematic variables (e.g., glide time and length, and step frequency) using simple and low-cost skate-mounted inertial measurement units (IMUs). Previous studies showed the method to be reliable, valid, and accurate in estimating the temporal and spatial parameters of forward skating in ice hockey [11,14,15]. In a recent study by Pojskic, et al. [10], the spatiotemporal parameters obtained by two skate-mounted IMUs were reliable and sensitive measures of sprint skating in male bandy players. Moreover, the metrics were useful in providing independent information for the different skating phases (i.e., the acceleration and maximal speed phase) and in predicting the maximal skating velocity with longer gliding and more frequent steps as the most significant determinants. Based on this recent study, it would be valuable to analyse whether sex and skating phases modulate the strength of the correlation between maximal velocity and the kinematic parameters.

Therefore, the main purpose of this study was to investigate differences between male and female bandy players in spatiotemporal variables during acceleration and maximum sprint skating (i.e., maximal speed phase). The secondary aim was to separately analyse sex differences in obtained correlations between maximal speed achieved in different skating phases and spatiotemporal metrics. Based on the above-mentioned studies, it was hypothesised that men would accelerate faster and reach higher maximum speeds than women due to shorter glide, double support times, longer steps, and double support length. Moreover, we hypothesised that the strength of the correlations would be skating phase and sex dependent.

2. Materials and Methods

2.1. Participants

A group of 74 female (age: 18.9 ± 4.1 years; height: 1.67 ± 0.06 m; body mass: 63.2 ± 7.4 kg; leg length: 0.92 ± 0.04 m; training experience: 13.4 ± 3.9 yrs.; and 26 elite and 48 junior elite) and 111 male bandy players (age: 20.7 ± 5.0 years; height: 1.80 ± 0.05 m; body mass: 76.4 ± 8.4 kg; leg length: 1.00 ± 0.04 m; training experience: 13.8 ± 5.0 yrs.; and 47 elite and 66 junior elite players) volunteered in the study. In total, 65 defenders, 73 midfielders, and 47 forwards participated. None of the players reported a history of neuromuscular disease or injuries in the previous six months. Two days preceding the experimental visit, players were asked to avoid sleep deprivation, refrain from high-intensity training, and avoid tobacco, alcohol, and caffeine. All participants were fully informed orally and in writing about the procedures and signed written consent (and by their parents when under 18 years of age) before participation. The study was conducted following the latest revision of the Declaration of Helsinki and current ethical regulations for research and was granted by the Swedish Ethical Review Authority (no. 2022-01550-01; approval data: 19 April 2022).

2.2. Procedure

Between January and March 2023, all testing was carried out at an indoor facility to eliminate the effect of weather conditions on results. Sixty minutes before testing the sprint-skating profile, leg length, height, and body mass were measured in each player. The leg length was measured as the distance from the anterior superior iliac spine to the medial malleolus to the nearest 5 mm [16]. Body mass was measured using a calibrated scale, and body length was measured using a tape measure attached to the wall. Afterwards, three players at a time performed their preferred warm-up for 15 min wearing their regular training gear to be ready to test 80 m maximal linear sprint skating. After their warm-up, the participants sat on a chair, and the testing procedure was described. A 16 g inertial measurement unit (IMU) integrated with a 3-axis gyroscope attached on top of the skate of each foot with a maximal measuring range of 2000°s⁻¹ ± 3% (Ergotest Technology AS, Statthelle, Norway). The sampling rate of the IMU was 200 Hz, but up-sampled to 1000 Hz before processing the step detection algorithm. After five minutes of rest, each participant performed two maximal 80 m sprint skating attempts with their club in their hand (to have the same condition as in competition), starting with the club behind the starting line.

2.3. Measurements

At the start and at every 10 m (10–80 m), pairs of photocells (Egotest AS, Statthelle, Norway) were placed at 1.2 m height to measure times at these distances and to calculate the average velocities and accelerations per 10 m. The participants started 0.5 m behind the first pair of photocells and had to skate for 90 m to avoid quitting too early before the finish line. All participants were encouraged to skate over 90 m in a straight line as fast as possible. Each player repeated the same procedure for two attempts, and only the best time taken to cover the 80 m distance in the sprint skating test was used in the data analysis. A rest period of five minutes was provided between attempts. The IMUs were synchronised with the photocells as part of the MUSCLELAB™ v10.57 system (Ergotest Technology AS, Statthelle, Norway). This made it possible to identify spatiotemporal skating stride variables automatically for each skating step using the software (Ergotest Technology AS, Statthelle, Norway) [12,14,15]. A skating stride is biphasic, consisting of support and swing phases. The support phase can be further divided into single and double support phases [17]. Propulsion occurs during both double and single support during the outward rotation of the thigh, coinciding with the initial extension of the hip and knee [18]. Glide time is the sum of double and single support time, and glide length is the total distance covered by both double and single support. Step length was defined as the distance covered at a single support, and the step frequency was calculated from the time elapsed between left and right foot touchdowns. By dividing the total distance covered in each step divided by the athlete’s leg length, the index glide length/leg length was calculated. The variables were analysed as (1) the average over 80 m, (2) over the first six steps (acceleration), and (3) over six steps at maximal velocity (i.e., the maximal speed phase) for male and female players as most acceleration occurs in the first 6 steps [10].

2.4. Statistical Analysis

The Shapiro–Wilk normality test was used to investigate if the data showed a normal data distribution. Homoscedasticity of variances was tested using Levene’s test. Since all the variables presented a normal distribution and homogeneity of variance, parametric tests were performed. The mean values and standard deviations were extracted from the descriptive statistics. To investigate the development of the average velocities and accelerations per 10 m between sexes, a 2 (sex: men, women) by 8 (every 10 m, repeated measures) analysis of variance (ANOVA) was performed with a Holm–Bonferroni post hoc test to identify changes in velocity ad acceleration over time. To compare the sprint skating profile and spatiotemporal skating variables during an entire run, acceleration, and steady state phases of male and female bandy players, an independent sampled t-test (sex: men, women) was used. The effect size was evaluated with Cohen’s d, where 0.2 < d ≤ 0.2 constituted a small effect, 0.5 < d ≤ 0.8 a medium effect, 0.8 < d ≤ 1.2 a large effect, and d > 1.2 a very large effect [19]. To investigate the correlations between skating velocity and step spatiotemporal skating variables, Pearson’s correlation was used per sex. Threshold values for the correlation coefficient interpretation as an effect size were 0.1–0.29 (trivial), 0.3–0.49 (moderate), 0.5–0.69 (strong), and 0.7–0.9 (very strong) [20]. All analyses were performed using JASP version 0.17.3 (JASP Team, Amsterdam, The Netherlands). Statistical significance was set at p < 0.05.

3. Results

The men were significantly older, taller, heavier and with longer leg length (p < 0.001) than the women, but in training experience, no significant differences were observed (p = 0.65,

Table 1. Anthropometrics and training years of male and female bandy players.

	Men	Women
Age (yrs.)	20.7 ± 5.0 *	18.9 ± 4.1
Body mass (kg)	1.80 ± 0.05 *	63.2 ± 7.4
Height (m)	76.4 ± 8.4 *	1.67 ± 0.06
Leg length	1.00 ± 0.04 *	0.92 ± 0.04
Training experience (yrs.)	13.8 ± 5.0	13.4 ± 3.9

* Indicates a significant difference between the sexes on a p < 0.05 level.

).

The development of skating acceleration and velocity was significantly different between men and women, as indicated by the significant group (F_(1,183) = 324, p < 0.001, d ≥ 3.09) and interaction effects (F_(1,1281) = 28.9, p < 0.001, d ≥ 0.92). Post hoc comparison revealed that every 10 m, the velocity increased while acceleration decreased, but that the men had a higher acceleration at almost every 10 m split, thereby reaching a higher velocity at every 10 m (Figure 1).

Men had a significantly higher skating velocity over the whole distance during the first six steps and at maximal velocity (t ≥ 2.29, p ≤ 0.029, d ≥ 0.45, Figure 2).

When comparing the skating spatiotemporal skating variables between sexes, gliding and double support time was significantly shorter in men compared with women in all phases and whole distance (t ≥ 3.1, p ≤ 0.001, d ≥ 0.63, Figure 3).

However, when comparing skating lengths between the sexes, a longer step length in each phase was found (t ≥ 2.01, p ≤ 0.048, d ≥ 0.40), while gliding length was only significantly longer when averaged over the whole distance and during maximal velocity for men compared to women (t ≥ 2.55, p ≤ 0.012, d ≥ 0.51) and no significant differences in double support length was found between the sexes for any of the phases (t ≤ 0.57, p ≥ 0.57, d ≤ 0.12, Figure 4).

Furthermore, step frequency and index glide length/leg length were significantly higher in men compared to women when analysed over the whole distance and during acceleration (t ≥ 2.19, p ≤ 0.031, d ≥ 0.44) and not during the maximal velocity phase (t ≤ 0.61, p ≥ 0.54, d ≤ 0.12, Figure 5).

Significant correlations were found between skating velocity and kinematic parameters. However, most were trivial to moderate, while glide and step length, and index revealed some strong correlations during acceleration (

Table 2. Correlations between velocity and the different kinematic parameters for acceleration, steady state phases, and the entire 80 m specified per sex.

	Average Over 80 m		Acceleration Phase		Maximal Speed Phase
	Women	Men	Women	Men	Women	Men
Glide time	−0.45 *	−0.21 *	−0.08	0.24 *	−0.26	−0.14
Double support time	−0.35 *	−0.13	−0.14	0.13	−0.11	−0.07
Glide length	0.20	0.24 *	0.52 *	0.60 *	0.33 *	0.28 *
Double support length	−0.06	0.04	0.15	0.30	0.11	0.08
Step length	0.30 *	0.34 *	0.62 *	0.70 *	0.36 *	0.36 *
Step frequency	0.29	0.20 *	0.09	−0.25 *	0.00	0.04
Index	0.22	0.23 *	0.49 *	0.54 *	0.33 *	0.29 *

* Indicates a significant correlation between velocity and this kinematic parameter on a p < 0.05 level.

).

4. Discussion

This study is the first to investigate the kinematic differences between men and women bandy players of linear skating sprint, explicitly focusing on the acceleration and maximal speed phases. The main findings were that men had a higher velocity and acceleration in all phases. Furthermore, they had shorter glide and double support times, longer step lengths, and higher step frequencies than women. In addition, glide and step length correlated the strongest with skating acceleration.

The male players exhibited better acceleration ability and achieved higher velocity at each measured split distance over the 80 m skating course than the female players, which corroborates recent studies that examined the sprint skating profile of competitive male and female bandy [5,6] and ice hockey players [7,9,21,22,23]. In general, the advanced skating sprint ability in male players was accompanied by shorter glide and double support times, longer step length, and higher step frequency than those of the female players. Interestingly, the power of some spatiotemporal metrics to discriminate the groups was shown to be skating-phase dependent. This corroborates previous ice hockey and bandy studies, showing the importance of analysing spatiotemporal sprint skating variables separately in the acceleration and maximal velocity phases [10,11,12].

Specifically, male players had a higher step frequency, but only in the acceleration phase. Conversely, the gliding length was only a discriminating factor in the maximal speed phase, with male players having a longer gliding length. However, when the glide length was normalised for the leg length, it also showed the power to discriminate between the sexes. Given that the glide length was the same in both groups but that men had a longer step length during the acceleration, it is logical to conclude that the smaller glide-to-leg length ratio was due to longer legs in men (i.e., 1.00 ± 0.04 m vs. 0.92 ± 0.04 m, respectively). The longer legs could enable male players to take longer steps and, therefore, to cover a longer distance. Considering that it was accompanied by a higher step frequency, shorter glide time and double support time, it was expected that male players could accelerate faster and reach a higher speed, that is to say, to skate over the same distance in a shorter time. This corroborates several studies that showed elite ice hockey players having greater stride rates than their lower-league playing counterparts [18,24,25].

At maximal speed, male players showed a longer glide and step length accompanied by shorter gliding and double support times. In contrast to the acceleration phase, the step frequency did not differ significantly between the groups at maximal speed. In other words, the more critical discriminating factors were gliding length and time than step frequency during the maximal speed phase. This means that male players achieved higher speeds due to the longer gliding length while maintaining the same number of strides as female players. It seems that male players obtain more benefits from the low ice friction that enables them to glide longer when transitioning from running to a gliding-like technique [26,27].

The observed sex dissimilarities in the skating spatiotemporal metrics (i.e., skating technique) and performance can be explained by multiple well-documented biological differences that could affect skating biomechanics and, thus, the push-off force production. For instance, the male players were taller (approximately 12–15 cm), heavier (approximately 18–22 kg), and had longer legs (approximately 8–12 cm) than the female players, which is in line with previous studies published on bandy [5,6] and ice hockey [7,21,22,28,29]. As theorised before, the longer legs in male players could enable longer step and stride length, which is especially important in the maximal speed phase. Given that velocity is the product of stride frequency and length, it is logical to conclude that longer legs could positively affect skating velocity. It seems that maximum velocity at maximal speed is more dependent on glide length and time than the number of strides in bandy players, which is, to a certain degree, contrary to studies on ice hockey players [17,24]. This can be additionally supported by the non-existing correlation between the maximal speed reached and step frequency in male and female players, with correlation coefficients as low as r = 0.04 and r = 0.00, respectively.

These longer step and glide lengths are probably caused by the ability to produce higher forward horizontal force. Biomechanically, this can be achieved by simply more knee flexion, which would be beneficial in two ways for producing advanced horizontal propulsive force. First, this could lead to a longer extension length and, therefore, to a longer path of applied force and, subsequently, higher skating speeds [18]. Second, the greater knee bend increases the contribution of the horizontal and decreases the contribution of the vertical force applied to the ice. In that way, the size and direction of the resultant propulsion skating force are flattened and, therefore, enhanced in the forward direction, which provides a skater with the mechanical prerequisites to accelerate faster. Third, the longer legs in men could provide a longer propulsive phase (i.e., time when force is applied), which enables muscles to generate and reach higher force output. This was observed in ice hockey by Shell, et al. [7] and Budarick, et al. [9], who found that males’ advanced skating acceleration and speed were accompanied by greater hip abduction and knee flexion at ice contact.

The other important factor in larger horizontal force production is more muscle mass in male players. This, together with a lower body fat mass, provides a higher relative strength and power and, thus, the advanced ability of male players to overcome body inertia, accelerate rapidly, and achieve higher skating speeds [5,6]. In the current study, the male players were significantly heavier (approximately 17–23 kg). Although we did not measure the body composition, based on the previous studies on sex differences, we can assume that male bandy players in our sample had a higher lean body mass and lower fat mass than female players. Men typically have 18–22 kg more muscle mass and 3–6 kg less fat mass than women [8] and lower fat percentage and higher lean body mass than female ice hockey players [21,22]. These sex differences in anthropometrics result in higher absolute and relative isokinetic strength of the knee flexors and extensors and better lower- and upper-body power, sprint, agility, and jump performance in male ice hockey players [7,21,22,23] compared to female players. Thereby, the well-documented biological dissimilarities also provide advanced biomechanical and performance-related prerequisites for male players to more efficiently produce and apply push-off force to the ice and propel themselves forward faster. Previous studies on ice hockey have shown that this can be accomplished through increased range of motion (ROM), strong activation of the hip flexors, and relaxation of the opposing muscles, such as the gluteus maximus [11,18,30,31]. Additionally, performing fast and extensive hip abductions results in a quicker swing phase, which leads to a higher step frequency and a longer glide [11,18,27,30,31].

As already shown to be important between the sexes, only the glide length and step length revealed some strong correlations during the acceleration phase, while all others were trivial to moderate. Interestingly, in male players, the step frequency showed a weak but negative correlation with acceleration ability (r = −0.25), which is contrary to a recent study in bandy [10] and earlier studies in ice hockey [12,25]. The observed discrepancies may be due to methodological differences that could affect the obtained correlations. Specifically, Pojskic, et al. [10] showed step frequency as a single kinematic predictor of acceleration ability in elite bandy players. However, they analysed fewer elite male bandy players (n = 32) than the current study (n = 111). When a more heterogeneous sample was analysed, the strong positive contribution of step frequency was diminished, while the other spatiotemporal metrics were demonstrated to be stronger determinants of acceleration ability. Similarly, Renaud, et al. [25] and Stetter, et al. [12] demonstrated that faster hockey players had similar step lengths but shorter contact and double support times, along with a higher stride frequency during the skating start. This could be expected because the players’ ability to accelerate over short distances and make consecutive changes of direction speed is essential for playing ice hockey successfully [13]. Conversely, in bandy, the ability to reach and maintain maximal gliding speed is more important [5,6,10]. This is also supported by higher step frequency in ice hockey than bandy players during the skate start [7,10].

Moreover, the differences between the constructions of the skate blades can potentially explain a lower contribution from the stride frequency in bandy than in ice hockey skating acceleration. Notably, the blades in ice hockey are typically shorter (approximately 2–3 cm) than in bandy and sharpened in a way that creates two edges with a hollow (i.e., channel) going along the blade [32,33]. This enables the blade edge to be dug deeper into the ice, providing ice hockey players with the necessary conditions (i.e., better grip) for strong push-offs and, therefore, the fast accelerations and stop-and-go manoeuvres that are crucial in ice hockey [13]. In contrast, bandy skate blades are longer without a hollow, enabling flatter contact and less grip with the ice, which enables longer glides at higher speeds but prevents players from making rapid accelerations and sharp turns. Consequently, it seems that in bandy players, glide and step length are more important than step frequency for accelerating at higher rates, which means that the length of the path at which they apply force is more important than how frequently they apply it. The obtained negative correlation suggests that increased step frequency might compromise acceleration in male bandy players. This finding confronts the idea that advocates stride frequency as the most important factor for the advanced acceleration in ice hockey [17]. Therefore, the ‘running-like’ technique in bandy players appears less expressed and important than in ice hockey players during acceleration.

Similarly, glide and step lengths significantly correlated with the maximal velocity in the maximal speed phase (even though it has a moderate correlation), while correlation with step frequency did not exist. This is contrary to previous studies in ice hockey that showed the importance of high stride frequency in discriminating elite from sub-elite players in achieving maximal skating speed [9,24]. The importance of stride frequency in reaching maximal skating speed in ice hockey might be due to methodological discrepancies between the studies. Namely, Upjohn, et al. [24] conducted their study on a skating treadmill, which precludes the comparison of the results with the current study, while Budarick, et al. [9] tested ice hockey players only over 34 m. This probably prevented players from reaching maximum skating speed, as seen from the marginal acceleration at the finish line. However, our findings corroborate, to some extent, the previous study in a small cohort of elite male bandy players [10]. Specifically, glide length was the strongest predictor of maximal skating velocity at maximal speed, while step frequency was a two-times less significant predictor than in the acceleration phase. Based on the optimal trade-off between increased step frequency and glide and step lengths, increased skating velocity in bandy seems to favour the latter. Considering the successful gliding technique in ice hockey, longer glides and, therefore, a longer time for the concentric force development can be achieved through increased hip extension and abduction, and knee extension at each push-off [11,24,25,27,34]. It appears that faster bandy players adapted the skating technique to meet the game requirements, rules, and skate construction, performing longer steps and glides, which compensate for the lower step frequency. Specifically, tackling is not allowed in bandy. Moreover, it is played on longer ice rinks than ice hockey (i.e., 100–110 m compared to 60 m). Together, these two factors enable bandy players to accelerate for longer periods, glide for longer distances, and thus reach higher maximal speeds [5,6,10] than ice hockey players, who are more involved in short sprints and changes in direction speed activities [13].

Several limitations of the current study should be acknowledged. First, the body composition and physical capacities were not measured. Second, neither joint kinematics nor muscle activation were examined. Third, we did not analyse the direction and position of the skate blades at each push-off phase, which could provide some additional explanations for observed performance differences between sexes. Furthermore, the study included a heterogeneous sample of elite senior and junior bandy players, which might limit its potential to discover the most important kinematic determinants of skating performance in a homogeneous top elite sample.

5. Conclusions

This study is the first to examine the kinematic differences between male and female bandy players in linear skating sprints, focusing on the acceleration and maximal speed phases. Male players exhibited better acceleration ability and achieved higher maximal velocity than female players. Better acceleration in men was accompanied by shorter gliding time, longer step length, and higher step frequency. When skating at maximal speed, male players had a longer step length and gliding time and length than female players. Step frequency did not correlate with the skating speed achieved in the acceleration or maximal speed phases. On the other hand, glide and step lengths significantly correlated with skating velocity in both the acceleration and maximal speed phases. Based upon these present findings, we suggest that coaches and players should, for faster skating in bandy, focus on taking longer steps and gliding for longer, rather than more frequently. Hence, players should be encouraged to stay low and have more knee flexion to enable a longer extension length and, therefore, a longer path and more horizontal direction of applied force.