The study of the actions done by the players and team (i.e., match analysis) is common in soccer (Carling et al., 2006). The goal of team analysis is to give coaches useful information about their players, team, and opponents in order to design training and to prepare their matches (Hughes and Franks, 2004). From a research perspective, the analysis of game has focused on finding patterns in the game or performance indicators (Hughes, 2003; Dufour et al., 2017). For this purpose, the observational methodology is used to collect information in team sports, as it allows for the collection of multiple variables that interact in the sporting context (Anguera, 2003; Anguera and Hernández-Mendo, 2015; Fabra et al., 2018; Maneiro et al., 2018; Nadal et al., 2018). In this sense, observational instruments are used to collect information about the behaviors done by players and teams in order to register indicators to improve sports performance (Anguera et al., 2017, 2018; Maneiro et al., 2017; Moreno and Gómez-Ruano, 2017; Morillo et al., 2017; Serna Bardavío et al., 2017; Chacón-Moscoso et al., 2018). The rules and characteristics of the game of soccer, such other field sports, made that these behaviors are complex in nature, uncertainty, and multifactorial (Vilar et al., 2012). For that reason, most of the observational instruments are created ad hoc to solve the needs of the coaches or to answer a specific research problem. However, there is also possible to find in the literature generic and specific observational instruments, such the GAIP-Soccer, TSAP, SOF-5, SOFBAS, FUT-SAT or SoccerEye. These instruments have similarities in the analysis of the players and team behaviors, such technical actions and spatial and temporal analysis. However, most of these instruments are just focused on the field players’ actions and they do not integrate the goalkeeper actions and the contextual situations of the players’ actions with the ball. The current development of football requires that coaches, researchers and performance analysts use instruments that brings together all aspects of the game. These instruments are required not only to integrate the technical actions of the outfield players, but also to analyze aspects such as the influence of the goalkeeper on the offensive game, the numerical situations that occur and the incidence of the outfield area on the technical–tactical actions that occur.

Match analysis has been part of soccer for a several decades (Hughes and Franks, 2004). The introduction of technology, especially specialized software and video, has increased the used of players’ and team analysis by coaches and researchers. The first observation instruments were linked to specific software’s [e.g., Noldus or Sportcode (Hughes, 2003)]. The analysis of data quality is one of the bases on which the observational methodology is based. This is critical to collect precise, objective and reliable data about the players’ actions. In the 1990s, several generic observation instruments were designed and validated. Examples of these instruments are the Team Sports Assessment Procedure (Gréhaigne et al., 1997) and Game Performance Assessment Instrument (Oslin et al., 1998). These instruments were generic, but they can be adapted easily to the different team sports (Harvey et al., 2010). In the two last decades, several specific observation instruments have developed in soccer. An incomplete list of these instruments is: (a) SOF (currently in its fifth version), initially developed by Anguera et al. (2003); (b) SOFBAS (currently in its second version) initially developed by Castellano (2000), (c) FUT-SAT, developed by Da Costa et al. (2011); and (d) SoccerEye, developed by Barreira et al. (2012). There is also possible to find in the literature a large number of observation instruments generate ad hoc to solve specific research problems, such as set-plays, activity profiles and group behavior (Sarmento et al., 2017). The analysis of these instruments and the proposal of categorical cores found in the reviewed literature show similar patterns. They collect information related the start of the offensive phase, the progress of the ball possession (attack and defense player/team), and the end of the offensive phase. The information collected about these phases described the actions done by the players and the situation in which these actions are done (e.g., temporal and spatial patterns). Most of these instruments are focused on the study of the field players. Only studies focus on the specific analysis of the goalkeeper had analyzed the actions of this player (Sainz de Baranda et al., 2005; Sainz de Baranda et al., 2008; Abellán et al., 2017).

Reviewed instruments analyze isolated actions, not integrated in the action in their context or players involved. In the same way, the analysis of the game of soccer requires a valid and reliable instrument. The dynamics of the game needs an instrument that analyzes the continuity of the player’s actions with the ball, without isolating the actions of the game. The current tendencies of the game require the integration of the goalkeeper as a critical element that participated on the team’s offense and the defense. The dynamics of the game need an instrument that analyses the continuity of the game, without isolating game actions. In addition, the goalkeeper’s offensive actions and numerical situations need to be integrated, as his participation in offensive actions in today’s football has increased (Florin, 2009; Szwarc et al., 2010). The proposal of observation instrument developed in this paper tries to integrate categorical cores from previous research and instruments and to integrate the goalkeeper in the analysis of the game players’ and team patterns. This observational instrument will provide information about how the technical and tactical actions are realized and will help to provide more information related to the dynamic system that involves team sport confrontation (Glazier, 2010; Castellano et al., 2017; González-Espinosa et al., 2017). The purpose of this research was to design, validate, and test the reliability of an observation instrument to analyze the offensive technical and tactical individual ball actions in soccer.

The list of categorical cores, degree of openness and their definitions after the first and second stages of the observational instrument design is shown in Tables 1–3. In the design of the observational instrument, the categorical cores and degree of openness were selected using the categorical cores and degree of openness proposed by Moreno (2005), Sainz de Baranda et al. (2005), Anguera et al. (2011), García-López et al. (2013), Hernández-Mendo et al. (2014), and Anguera and Hernández-Mendo (2015).

In the third stage, the experts reviewed the observation instrument. The experts’ qualitative observations were related to the definitions of categorical cores. No categorical cores were eliminated by the experts after the experts’ evaluation. In the quantitative evaluation, all categorical cores had an average score >0.70 (Table 4). The lower Aiken’s V value found was 0.91, concretely in the categorical core “numerical situation with opponent goalkeeper.” In the fourth stage (observer training), the lowest intra and inter observer agreement coefficient observers was >0.96 for all the studied categorical cores (Cohen’s Kappa and interclass correlation coefficient); and the lowest intra and inter observer agreement coefficient observers was >0.97 for all the studied categorical cores (Kendall’s Tau B) (Table 5).

Finally, the analysis of generalizability (Tables 6, 7) shows in the first design a Generalizability coefficient (GC) equal to 1, this result shows a high reliability of the observers. In the second and third design, the results of the GC are equal to or very close to 0, these results indicate a high adjustment of the observation instrument and that its categories were exhaustive and mutually exclusive (E/ME) (Anguera and Hernández-Mendo, 2013). The highest percentage of variance (see Table 7) is found in the interaction [Criteria] [Categorical cores], being very low in the rest of the sources of variation.

The present paper describes the stages done to design, validate, and test the reliability and generalizability of the observational instrument. Considering the increasing number of observational studies that use “Ad hoc” instruments in football, the development of this instrument aims to provide the many observational studies that occur in football with a valid and reliable instrument that allows adequate data collection. In the same way, the developed instrument has the advantage of using open categories, as opposed to other studies that present closed categories, in which only the conducts carried out are registered. In addition, the designed instrument has the advantage that it takes into account the continuity of the game. The instrument registers the actions that happens since the players obtain ball possession until they lose it.

The fact of being able to record the continuity of the offensive game, the procedures by which it is made with the ball, the behaviors it manifests in possession of the ball and finally, the actions developed when it is released from the ball, is an advantage over other instruments that analyze and record actions in an isolated way, for example instruments that analyze goals (Reina-Gómez et al., 2010). In addition this instrument provides, in spite of analyzing the continuity of the offensive action, do not take into account categorical cores of relevance as the actions of the goalkeeper, the corporal zones used in the action, the existence of companions in support or the defensive lines that are surpassed in the offensive action (Sarmento et al., 2010).

The process had different stages, similar to the one followed in the development of observational instruments in other sports (Villarejo et al., 2014; Palao et al., 2015a,b). The first stage involved the review and analysis of the available observation instruments and match analysis literature. The analysis of the available material was focused in the design an instrument that allows obtaining information: (a) about the individual and collective game patterns, (b) about the players evolution in training and competition, (c) about the effect of manipulation of rules in small side-games or in competition, and (d) that guide the training design. After the analysis, the researchers established the categorical cores focus on the following aspects: the sequence of the ball possession (stat, develop, and end), description of the actions done by the player with the ball, description of the temporal, spatial, and context of the actions done with the ball, and inclusion of the goalkeeper in the analysis of the situation. The researchers tried to use in the design of the instrument categorical cores close to the criteria used by coaches with the goal to increase the applicability of the instrument. For that reason, the criteria used for the Spanish Soccer Federation to describe the technical and tactical actions with the ball in the coach training courses was followed (Moreno, 2005). Along with this criterion, it was also used the categorical cores proposed by researchers in other instruments and research studies (Sainz de Baranda et al., 2005; Anguera et al., 2011; García-López et al., 2013; Hernández-Mendo et al., 2014; Anguera and Hernández-Mendo, 2015). The instrument has as unit of analysis the ball possession and differentiates in the ball possession three moments (start, development of end), like other instruments in soccer (e.g., SOFBAS or SoccerEye). The pilot studied carried out by the researchers allowed reviewing the categorical cores and their definitions.

In the third stage of the process, the experts’ qualitative and quantitative revision and the use of Aiken’s V allowed for measuring the content validity of the items. The values of the quantitative evaluation were high for all the categorical cores (Vo ≥ 0.70) and above the minimum values proposed in the literature as a reference (Vo = 0.70) by Penfield and Giacobbi (2004). The qualitative evaluation helped specifically to clarify some of the aspects of the definitions of the different degree of openness of the categorical cores. The use of experts from research and coaching provided a more complete and holistic vision of the sport.

The level of intra-reliability, reached between observers after the observer training, showed the instrument is adequate with regard to reliability (Bakeman et al., 1997). The use of categorical cores and degree of openness well established in the area and the qualitative evaluation of the experts contributed to the high level of reliability. Finally, the results shown in the generalizability analysis made it possible to verify the high levels of validity, precision of the instrument and corroborate the high reliability indices of the observers (Blanco-Villaseñor et al., 2000; Hernández-Mendo et al., 2016).

Data obtained by the instrument can provide useful and applicable information to coaches in order to understand game patterns, face the competition and develop the training sessions. The structure of the instrument allows coaches and researchers to use it in its entirety or certain parts or criteria can be utilized to solve specific research problems.

The results showed that the observational instrument is valid and reliable for measuring the technical and tactical actions done in the offensive phase by the player and team with the ball possession. The instrument has some limitations. Only it was assessed the content validity of the instrument (expert evaluation), and the instrument is focused on the actions of the players and team which the ball possession. However, the instrument can provide coaches and research with information about the type of actions done by the players with the ball, their characteristics and in which context are done (distance, numerical situation) and the level of involvement of the field players and the goalkeeper. This information could allow coaches and research to establish the demands of the game and to create training plans that help players to be prepared for these demands.

This study respected the ethical principles established by the UNESCO Declaration on Bioethics and Human Rights. The parents or guardians of the players were informed of the study and gave their written consent in accordance with the Declaration of Helsinki. The study was approved by the Ethics Committee of University of Murcia (Spain) with ID 1944/2018.

All authors participated in the design, documentation, development, and writing of the manuscript. This paper was reviewed by all authors and all of them are responsible for its contents and providing they are responsible for the final version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.