Determinant Factors for Throwing in Competition in Male Elite Handball

: The aims of this study were to deﬁne determinant factors that affected throwing actions used to score goals in handball, and to study the effectiveness of throwing actions and their relationships with different factors during their use, such as: the throwing distance, type of arm build, type of throw, whether it is done while jumping or while supported, the last supporting foot and the previous step cycle, as well as their relationship with the ﬁnal score of the game in elite male handball competitions. A total of 1049 shots from 24 matches in the 2018/19/20 International Men’s Handball Championship were analysed. The results show that distance is relevant in the effectiveness of throwing, setting the build up, throwing while supported or jumping, the type of throwing, the type of steps cycle used and the supporting foot. This seems to indicate that the determining factor that affects a throw is distance.


Introduction
The throw used to score a goal in handball is defined as the most important action for success [1]. Determinant factors analysis of handball throwing is focused on variables such as: distance; throwing zone, type, angle or position; the result or final effectiveness of the total quantity of actions [2][3][4]; or factors of vital importance in the result of the throw, such as speed and accuracy [5][6][7][8][9][10]. In this line, recent research shows that a shorter throwing distance predetermines the effectiveness [2,7,11]. In addition, anthropometric characteristics, throwing velocity [12] and technical-tactical skills are also important for successful participation in elite-level handball [13]. In addition, various articles used the throwing variables as discriminant elements between the winning and losing teams [14][15][16][17][18].
Notational analysis allows us, from an ecological view, to analyse the competition [19,20]. In addition, it is necessary to understand the behaviour in these clashes similar to the interaction of dynamic and complex systems, which require the analysis of both individual and collective decision making, considering the context [21][22][23]. For this reason, many articles focus on match analysis in handball [8,14,24,25].
It is very important to focus on the determinant factors that affect competition throwing in order to focus on the technical learning process in handball. Therefore, we must answer the needs of coaches, who need to know which factors affect the performance of throws in elite handball. Given this context, the aims of the present study were to define determinant factors that affect the throws used to score goals in handball, and to study the effectiveness of throwing actions and their relationships with different factors during their use, such as: the throwing distance, type of arm build, type of throw, whether it is done while jumping or while supported, the last supporting foot and the previous steps cycle, as well as their relationship with the final score of the game in elite male handball competitions.

Materials and Methods
The observational methodology enabled data collection directly from the participants in the competition (notational analysis) [20], including analysis of recordings and sequences of the technical-tactical actions of the game in collective sports and its result (match analysis) [24,25]. The ethical and deontology principles were met in relation to the participants in the research, and the data obtained was managed according to the ethical principles of the Helsinki 2016 declaration and approved by the Ethics Committee of CEIC Aragón (CEICA) no 10/2021.

Participants
The sample was composed of male elite national teams of the European Championship in 2018, World Championship 2019 and Europe Championship 2020. The teams classified between the first and the fourth were selected. We observed 12 matches. We studied 24 team analyses in the finals, semi-finals and third and fourth place matches. A total of 180 players were analysed (mean age: 27.57 ± 4.3 years; body weight: 86.6 ± 9.9 kg; body height: 1.90 ± 12.2 m; training experience: 10.1 years; training work: 23.5 h per week).

Instruments
To analyse the determinant factors affecting throws in elite handball, an ad hoc observation system was designed [20]. It was composed of eight variables and 28 multidimensional categories ( Table 1). The computerization of the registry was conducted by the free and versatile software Lince v.1.4 (www.observesport.com, accessed on 24 September 2021) [26]. This program contributes to computerized monitoring procedures that speed up the registry process [27].

Variable Description
Jump/Stand − Jump throw: Throw done in the air phase of the jump when the player is not in contact with the ground.
− Stand throw: Throw done when the player is in contact with the ground with one of his feet.
Type of throw − Speed throw: Any other type of throwing that is not considered as skill throwing.
− Skill throw: The player uses some sort of high-level technique such as a screw (throwing with effect), a topspin throw (a throw in which the ball in its air path changes its speed) and parabolic throwing.
Step cycle − Zero step: A throw without using any step from the step cycle.
− One step: Throw done after taking a step.
− Two steps: Throw done after taking two steps.
− Three steps: Throw done after taking three steps.
− Flying: Throw done when the ball is caught in the air and thrown before it touches the ground.
− More than three steps: Throw done with more than three steps without being disciplined for this regulatory violation. − Two feet: Throw done with the last contact of the player with the ground being with both feet simultaneously.

Results
− Goal: A throw that is granted as a goal by the referees due to exceeding the net line.
− Out: A throw that is not touched by any player of the rival team and ends out of the net or hits the bars without being a goal.
− Blocked: A throw where the goalkeeper prevents the throw from ending up in the goal.
− Defence: Contact/action of the defender on the ball throw.

Data Validation
The validation instrument was validated by an expert panel formed by three sports sciences graduates and handball national coaches with research experience in observational methodology [28]. Training of observers was completed thanks to the creation of an observation manual in which variables and the observational process codes were defined [29]. All of the analyses were conducted in a 30-day period using the same tool and in the same space. Two randomly chosen matches were analysed, calculating the intra-observer internal consistency and reliability (at two different times) and inter-observer internal consistency and reliability. The internal consistency and the reliability thresholds, between 0 and 1 [30], were set at: for α (internal consistency), <0.50 unacceptable, 0.51-0.60 poor, 0.61-0.70 questionable, 0.71-0.80 acceptable, 0.81-0.90 good, and ≥0.91 excellent [31]; for ICC (reliability), ≤0.50 poor, 0.51-0.75 moderate, 0.76-0.90 good, and ≥0.91 excellent [32]; and for κ (reliability) <0.01 no agreement, 0.01-0.20 poor, 0.21-0.40 discrete/regular, 0.41-0.60 moderate, 0.61-0.80 good, and 0.81-1.00 very good [33]. The validation showed the internal consistency and reliability of the intra-observer means as 0.86, and inter-observer means as 0.91. Both can be considered good and very good.

Statistical Analysis
Basic statistical descriptors (frequency and percentages) were calculated for each game-related statistic using the match outcome (winning and losing teams). Normality was calculated for each variable using the Kolmogorov-Smirnov test. For the differences between the winning/losing teams, an ANOVA (one factor) test was applied. Additionally, the chi-square test was used to examine the relationship between the variables, where the direction of the relationship was identified using corrected standardized residuals (Zcorrected). A p-value < 0.05 was considered to be statistically significant. The statistical analysis was performed with the software package SPSS version 25.0 (SPSS Inc., Chicago, IL, USA).

Results
A total of 1049 throws were registered ( Table 2). Throws were most frequently made from 6 metres (44.7%). The most used throwing technique was the classic (93.2%). Most of the throws were made while jumping (77.1%). Most of the throws were at the maximum possible ball speed (95.7%). The most frequent throw was made with two steps (31%). The total efficacy was 62.2%. The analysis of variance (one-factor ANOVA) regarding variables of ranges with a normal distribution (Table 3) did not show significant differences between the average of the different championships and the quantity of throws (1.890 f; 0.176 sig) and their effectiveness (0.542 f; 0.590 sig). In the chi-square test analysis (Table 4), 13 significant relationships were found (p < 0.05). The results of the throws was related to distance (125.39, p < 0.001) and the type of throw used (12.62, p < 0.01). In addition, distance was related to arm build (92.84, p < 0.001), jump/support (459.45, p < 0.001), the type of throw (62.01, p < 0.001), the step cycle (564.80, p < 0.001), the supporting foot (26.91, p < 0.001) and match result (14.27, p < 0.05). Additionally, arm build was correlated with the jump/support variable (71.69, p < 0.001) and the supporting foot (33.49, p < 0.01). The jump/support variable was related to the step cycle (198.92, p < 0.001) and the supporting foot (20.32, p < 0.001). Finally, the supporting foot had correlation to the step cycle (56.74, p < 0.001). Between the result of the throw and distance, a positive and negative correlation exists in most of the categories of each variable ( Table 5). The following significant relationships were found between the result of a throw and certain variables: block and 9 m (z = 6.5); goal and 6 m (z = 5.7); out and 9 m (z = 4); save and 9 m (z = 3.9); between hip and arm position and 6.9 (z = 7.9); jump throw and 6 m (z = 13.8); winner teams and middle field zone throw (z = 2.9). Table 5. Residual z punctuations between throw distance and significant variables. Analysis of the step cycle (Table 6) shows the high z scores of throws from from 7 m and zero steps (z = 22.4) and supports the hypothesis that throws taken with zero steps are successful (z = 14.0). Table 6. Residual z punctuations step cycle and significant variables.
Step Cycle

Discussion
The aims of this research were to analyse the determinant factors that affected throws in handball, their effectiveness and their relationships between the variables that defined the results of throws in handball competitions of the elite male category. This study was framed by the study of match analysis [24,[34][35][36].
The results highlight that the effectiveness of a throw is predetermined by the throwing distance and the type of throw. Equally, relationships exist between the throwing distance and all the other variables.
Originally, we found a relationship between throwing distance and its result. Similarly, Avila (2018) [24], Blanco García (2012) [37] and Almeida et al. (2019) [8] observed that throwing from six metres has a positive relationship with goal achievement, compared to throwing from nine metres, where that relationship is negative [38].
Furthermore, significant relationships have been identified with throwing distance and the result of a throw. A throw, when performed from the middle distance (9 m) and between lines (6-9 m) needs to be strong and powerful so that it can take the defence by surprise [1,39,40].
Equally, the relationship between the throwing distance and type of throw is brought to light. This fact is also proven in other team sports [34][35][36]41]. Additionally, the throw with the classic technique is the most used in throws from distances of six metres. This type of throw allows for greater manoeuvrability and variability in the launch location [42]. Furthermore, throwing while jumping is significantly more repeated and effective in throws of short distances. This is because the player has more time to observe the goalkeeper and make the right decision [43].
The step cycle offers interesting positive relationships with throwing distance. The use of no steps is found mainly in throws made between lines (6-9 m), utilizing the factor of surprise motivated by the closeness of the defensive line. A nearby throw (6 m) requires a great manoeuvre speed in little space, which is why it is commonly used with a one-step throw. A two-step throw has a greater association with the distance throw (9 m), lending a higher inertness to the throwing action by the tagging actions of the defensive line [44].
Step cycle management training related to throwing is one of the actions that cannot be missed during the formation of a handball player [39,40].
Therefore, it is important that handball coaches give importance to the variability of throwing work.

Conclusions
In terms of the descriptive analysis from the last three years, in male selection championships, the quantity of throwing in semi-finals and finals has a slight downward trend in comparison to the prior championship.
The relationships observed in the research suggest that the result of a throw depends on the distance and type of throw. The distance is relevant to most parts of a throw. Therefore, the search for effective throwing actions motivates teams to tailor their training sessions to optimize the search for goal-scoring situations close to the handball goal. It is important to increase the training time for throwing in handball to improve throw variability.
A correct approach in the learning process is key to the education of handball players. Taking advantage of the dynamic richness step cycles offer is a skill that is essential in order to reach the highest level of offensive performance in handball, as the step cycle is an essential feature of handball that helps finish a throwing action with effectiveness.