Analyzing coordinated group behavior through role-sharing: a pilot study in female 3-on-3 basketball with practical application

A group often shares a common goal and accomplishes a task that is difficult to complete alone by distributing roles. In such coordination, the non-verbal behavior among three or more members complicates the explanation of the mechanism due to complex and dynamic interactions. In cognitive science, a crucial role is indicated: to intervene moderately with others and adjust the whole balance without interrupting their main smooth interactions, using an experimental task. The findings suggest that resilient helping actions in the third role support coordination. These actions are similar to off-ball movements in team sports, which involve an on-ball player and have recently been the focus of sports science because their characteristics are not represented in common statistical data, such as a shooting success rate. Hence, a new perspective for discussing coordination has emerged, as existing theories, such as synchronization—where movements between players are spontaneously matched and organized—cannot explain the mentioned role. However, there is a lack of investigation and discussion regarding whether these findings are applicable to real-world activities. Therefore, this study applied the experimental findings to the field of sports. We developed a 3-on-3 basketball game in which the offensive role of intervention decision and adjustment is key for winning and introduced it to the practice of a female university team as a pilot study. Participants repeatedly engaged in the mini-game, and the playing was compared before and after receiving tips for this role. Consequently, in the bins of the relatively large distance between the participant required to the relevant role and each defensive player, the frequencies after receiving these tips were significantly higher. Furthermore, the winning rate on the offensive team improved temporarily; however, the effects were not maintained. These suggest that spacing skill, which maintains reasonable distances from the other players, creates favorable situations for coordination. This study may bridge the gap between controlled experiments and real-world applications and make an educational contribution; it may recommend practice design for the acquisition of spacing skills related to the crucial role.

1 Introduction

A group often shares a common goal and accomplishes a task that is difficult to complete alone by distributing roles among its members. This is explained by common opinions or suggested through experiments and simulations in many previous studies [e.g., (1–4)]. This idea is encapsulated by the saying “Two heads are better than one” and is defined as planned coordination (coordination) in psychology and cognitive science (5). For coordination, members share some roles considering the workloads and task specifications. Distributed cognition theory explains that an overall group function works through interactions among subsystems and between subsystems and environments, in which each subsystem plays a role (6, 7). The internal and external resources are properly distributed, and each member's workload decreases. Furthermore, cognitive interactions among roles based on different perspectives lead to problem-solving and the discovery of various strategies [e.g., (8–12)]. Team sports, in particular, are typical examples of the benefits of role-sharing by multiple on- and off-ball players [e.g., (13–16)]. This is required to achieve a common goal and high performance under rules, time constraints, and defensive pressures. Team sports often involve three or more members and aim to achieve a common goal through complex and dynamic interactions. Complexity indicates that explaining the coordination mechanism is more complicated than for a pair because relationships among members diversify, such as role-sharing, and are not interpreted by one-dimensional behaviors such as leader-follower and approach-avoidance (17, 18). Dynamics suggests non-verbal and time-series features that develop strongly over a short period, such as body movements (19, 20). Cognitive science focuses mainly on higher-order information processing, and coordinated behaviors among three or more members are not fully investigated. Considering that physical interactions are primitive and that a group of three or more members is often observed in real-world activities, it is important for cognitive science, which discusses social intelligence, to investigate complex and dynamic coordination (17).

This study focuses on a crucial role in the coordinated behavior of a triad through role-sharing, which is indicated using an experimental task in a cognitive science study (21). We aim to bridge a gap between experimental findings and real-world applications and contribute to the understanding and development of coordination mechanisms.

In related works, a sports science study recorded 5-on-5 basketball games and identified the coordinated defensive structures of role-sharing according to emergent situations by a top-level male university team in Japan (1). Recently, machine learning studies have predicted shooting success rates and evaluated coordination leading to goals by classifying and extracting features of physical interaction structures with role-sharing using datasets in basketball and soccer [e.g., (22–24)]. Furthermore, network science studies have identified indices of centrality in network models where nodes and links are regarded as players and pass in soccer, representing team-specific strategies and key players related to robust coordination [e.g., (25–27)]. Meanwhile, collective behavior studies observed coordinated hunting and found that the role-sharing of chasing and blocking naturally emerged [e.g., (28, 29)]. Additionally, multi-agent simulation using deep reinforcement learning can replicate this process. In the simulation, agents perceive and observe the environment, including the prey, and learn the optimal behaviors to maximize their rewards (30). The findings in sports science, machine learning, network science, and collective behavior suggest that role-sharing creates favorable situations for coordination. In principle, these studies aim to understand the physical interaction structures in coordination. However, the information processing underlying these characteristics has not been fully discussed, and educational applications have not been considered.

Notably, although not competitive, a previous cognitive science study investigated a crucial role in the coordinated behavior of a triad through role-sharing using an experimental task (21). This indicates that the role of intervening moderately with other roles and adjusting the whole balance was related to high task performance. Such resilient helping actions are an important factor for successful defensive coordination in team sports and effective team building in business and military organizations, not only in experimental tasks [e.g., (1, 31–34)]. This role is also required not to interrupt their main smooth interactions. The third role needed to decide whether to intervene according to the situation. A new perspective for discussing coordination has emerged because existing theories in cognitive science and sports science, such as synchronization [e.g., (35–38)]—where movements between players are spontaneously matched and organized—cannot explain this role. The findings suggest that adjustment to create favorable situations without interrupting others supports coordination. This concept is similar to off-ball movements in team sports involving an on-ball player. For example, in basketball, when an on-ball player is surrounded by defensive opponents, another offensive player approaches and directly receives a pass for helping. Off-ball movements in basketball and soccer have recently been the focus of sports science, as shown by distances between players and those with the goal. These contain valuable information on coordination and are not represented in common statistical data, such as shooting success rate [e.g., (13, 16, 39, 40)]. The previous study (21) confirmed the crucial role in coordination; however, there is a lack of investigation and discussion regarding whether these findings are applicable to real-world activities.

Therefore, we applied the experimental findings to the field of sports as a pilot study. This study focused on team sports that must achieve a common goal and high performance within some constraints. We used 3-on-3 basketball in which the coordination by off-ball players is essential for winning, as mentioned above. Furthermore, it is easy to record group behavior because of the relatively small number of players (three on each team) and the court size. Additionally, it has recently attracted worldwide attention, as evidenced by its inclusion as an official Olympic event. Meanwhile, few studies have used 3-on-3 basketball to discuss coordination in terms of cognitive science. The purpose of our study was to investigate the influence of the role of intervention decision and adjustment on coordination in 3-on-3 basketball. We developed a mini-game in which the relevant offensive role is key and introduced it to the practice of a female university team. The players repeatedly engaged in the game. After the first half, the offensive team received tips on coordination focusing on this role. The team performance and playing related to the relevant role were quantitatively compared before and after receiving these tips. According to the purpose, this study has set the following hypothesis: after receiving the tips, the team performance improves, and the appropriate role executions are observed.

Our study connects cognitive science and sports science because this pilot study includes an investigation of information processing underlying complex and dynamic coordination, which has not yet been fully discussed. The quantitative analysis may recommend practice design on playing that helps an on-ball player and creates favorable situations in the fields of sports. Bridging the gap between controlled experiments and real-world applications is a challenging endeavor. These findings may also offer implications for further developing 3-on-3 basketball itself. Next, we explain the details of this practice.

2 Methods

2.1 Participants

Six female students on the university basketball team, of which the second author is the head coach, participated in this practice. They usually play 5-on-5 in the official game. This team was affiliated with the third division of the Tokai area league in Japan, practiced regularly for about 2 h each time three or four times a week, and played approximately 30 games per year including practice matches. The participants included five regular players and a sixth man. The sixth man means a first substitute player. Their averages of age, height, and basketball experience were 19.17 age (SD = 0.90), 165.08 cm (SD = 5.88), and 10.58 years (SD = 2.70), respectively.

The participants were divided into the offensive and defensive teams. The head coach conducted the team compositions to make their abilities competitive based on their profile data of age, dominant hand, height, basketball experience, and current and previous positions (see the details on the offensive and defensive teams in Supplementary Materials). The second author has 25 years of coaching experience and holds a certified B-level coaching license from the Japan Basketball Association. The record includes leading a university team to promotion to the first division of the Tokai area league and participation in national tournaments in Japan. The participants on each team were fixed. This study statistically compared the offensive team performance and playing related to the role of intervention decision and adjustment before and after receiving the tips of this role. Hence, we planned this practice based on the policy of eliminating factors other than these tips as much as possible. Meanwhile, it should be kept in mind that we applied the experimental findings to the field of sports as a pilot study.

2.2 Informed consent

We explained how we would video-record and collect data. When explaining, not all the participants were informed of the details of the procedures in this practice. They were not made aware of the playing related to the role of intervention decision and adjustment, receiving the tips of this role to the offensive team, and comparing the offensive team performance and playing before and after receiving these tips. The procedures were also based on the policy of eliminating factors other than the tips. Written informed consent was obtained from all the participants. This study was approved by the ethics and safety committee of Shizuoka University and Tokoha University, to which the first author and the participants were affiliated. Our study was conducted following these regulations. According to the informed consent, not all images in the manuscript contain individual identifiers.

2.3 Developed 3-on-3 basketball game

Figure 1 shows the diagram of the developed mini-game in which the circles and squares represent the offensive and defensive players, respectively. Players from #1 to #3 on each team are the same participants in all the games (trials). Offensive #3 is required to play the role of intervention decision and adjustment focused on in this study. The winning condition for the offensive team is that within a 15 s time limit, someone draws a defensive player and a shooter is unmarked in open space. Conventionally, the defensive team must play without satisfying it. To investigate coordination, the above condition is defined independently of individual skills. The team performance reported in the Results section is consistent with this definition.

Figure 1

Figure 1. Diagram of the developed 3-on-3 basketball game in this study. In this mini-game, the circles and squares represent the offensive and defensive players, respectively. The winning condition for the offensive team is that within a 15 s time limit, someone draws a defensive player and a shooter is unmarked in open space. At the start of the game, the ball handler of #1 is fixed, and the player begins without directly passing to #3. The starting position of each offensive player is assigned (see the Developed 3-on-3 basketball game section to confirm the other rules). For the offensive team to win, #3 is a key player; it is required to intervene with the other players, adjust the two-man game, and create favorable situations, such as (1) handing the ball off directly with #1 or #2 (“hand-off”), (2) working as a wall to interrupt the defensive player (“pick and roll”), and (3) moving to a corner area and taking a three-point shot. Conversely, it is crucial not to interrupt #1 and #2 without forcibly intervening according to the situation.

In the mini-game, we set some rules to guide the same situation at the start of the game and compare the playing before and after giving the tips about the role of intervention decision and adjustment. The ball handler #1 is fixed, and the player begins the game without directly passing to #3. The starting position of each offensive player is assigned to maintain a certain distance to observe the coordination process in which this role involves the other players. On the defensive team, #1 and #2 confront offensive #1 and #2 and their starting positions are voluntary. The start position of defensive #3 is also voluntary. Just before the game, the two-man game is randomized on the right (Figure 1) or left side of the basket goal by instructions to eliminate the influence of the dominant hand. The starting positions on the offensive team are called the two-guard and also used in a 5-on-5 game. For the offensive team to win, it is important that #3 intervenes with the other players, adjusts the two-man game, and creates favorable situations, such as (1) handing the ball off directly with #1 or #2 (“hand-off”), (2) working as a wall to interrupt the defensive player (“pick and roll”), and (3) moving to a corner area and taking a three-point shot. Conversely, it is also key not to interrupt #1 and #2 without forcibly intervening according to the situation. The offensive coordination required in this game is fundamental. However, the head coach mentioned that the participants could not implement such interaction in official games at that time.

2.4 Environment and procedures

The upper part of Figure 2 shows the environment of the mini-game conducted in the university gymnasium. The court area, including the vertical endline and the area under the basket goal (“paint area”), followed the official size of 3-on-3 and 5-on-5 (41, 42), except for the horizontal sideline. A digital countdown timer was used to watch the remaining time (Mollten Corp., UX0110). The participants regularly practiced on this court size in the gymnasium. Hence, factors such as the court area, lighting, and flooring were unlikely to negatively influence the offensive team’s performance and group behavior. The mini-games were recorded from a bird's-eye view on the stage using only one video camera, as shown in the lower part of Figure 2 (Sony Corp., HDR-CX680). As they played all the trials within the camera's field of view, there were no issues with tracking their positions by image processing technology as below. We preliminarily prepared another video camera; however, no equipment problems occurred during this practice. Consequently, these recordings were excluded from this analysis.

Figure 2

Figure 2. Environment in this practice. We conducted the developed 3-on-3 basketball game in the university gymnasium. The court area, including the vertical endline and the area under the basket goal (“paint area”), follows the official size in 3-on-3 and 5-on-5 (41, 42), except for the horizontal sideline. All the images in the mini-games were recorded from a bird's-eye view on the stage using only one video camera by the first author. After receiving the tips about the role of intervention decision and adjustment, #3 required for this role, wearing a pink bib as shown in Figure 1, moves to a corner area while maintaining a reasonable distance from each defensive player and aims to take a three-point shot, as shown in the lower images. We explained how we would video-record and collect data. Written informed consent was obtained from all the participants. According to the informed consent, all the images are shown while blurring some parts to avoid identifying individuals.

Regarding the procedures, the experimenter briefly announced the practice schedule (see these details in Supplementary Materials). Subsequently, the participants warmed up and received colored bibs for individual identification (Figure 1). The experimenter then explained the team assignments and mini-game rules. At this point, the first author did not instruct the offensive team on the role of intervention decision and adjustment and playing. The three practice trials, including a trial by the timer error, were conducted and they could confirm the rules. Subsequently, in the first half of the pretest, three sessions comprising seven trials per session were conducted for a total of 21 trials. The interval between the trials was approximately 30 s, with approximately a minute between the sessions. The university team regularly played a 5-on-5 game with each quarter for 10 min with the 2 min intervals for a total of 40 min across the four quarters. In comparison, the total of playtime in this practice was a maximum of approximately 10 min. Hence, fatigue was unlikely to negatively influence the offensive team performance and group behavior. Furthermore, we carefully checked the participants just before restarting the mini-game to ensure no fatigue. After the first half, a guest coach who understood the purpose of this study gave each team tips based on the observations of the games. The coach was a professional player in Japan and is currently the head coach of a junior youth team. In this practice, giving tips that reflected the essence of our study to the offensive team was indispensable. Thus, we asked the guest who is superior to coaching to do so. The tips for the offensive team focused on coordination related to the role of intervention decision and adjustment. Meanwhile, the defensive team received the general tips on help defense, excluding participant- and game-specific strategies. Table 1 presents the details of the tips given to each team. While one team received them, another waited at a distance to prevent information gathering. Generally, it is common for a team to keep strategies hidden from the opponent. The purpose of this study was to investigate the influence of the tips on the offensive team performance and playing related to this role. Hence, we planned this practice based on the policy of eliminating factors other than giving offensive tips; the mentioned procedure was appropriate. Subsequently, the second half of the posttest was conducted similar to the first one.

Table 1

Table 1. Tips to the offensive and defensive teams by the guest coach.

3 Results

3.1 Team performance

This study counted the winning numbers of the mini-games for the offensive team to compare them between the first and second halves. Those indicated in the first half for a total of 21 trials from Sessions 1 to 3 and in the second one from Sessions 4 to 6 were 9 and 11, respectively. Table 2 represents a cross-tabulation table of the event and performance. A chi-square test indicated no significant relationship between them, and the effect size φ was small (${\chi }^2(1)=0.095,p=0.757,\phi =0.048$). Meanwhile, Figure 3 shows the progression of the winning rates on the offensive team over the sessions. The rates of Sessions 1 and 2 in the first half were at the chance level, indicating four wins in the seven trials; thereafter, it decreased in Session 3. Notably, in the second half, it recovered in Session 4 and drastically improved in Session 5, indicating six wins; however, it significantly decreased in Session 6. It indicated that after receiving the tips about the role of intervention decision and adjustment, the offensive team performance improved temporarily but was not maintained.

Table 2

Table 2. Cross-tabulation table of the event and offensive team performance.

Figure 3

Figure 3. Progression of the winning rates on the offensive team over the sessions. Each session consists of seven trials. After Session 4, given the tips about the role of intervention decision and adjustment, the team performance improves temporarily but is not maintained.

In summary, giving the tips about the role of intervention decision and adjustment did not fully influence the offensive team’s performance. The main reason for the decrease in the winning rate in the final session was that the defensive team had established an effective countermeasure, as reported by the participants on the defensive team (see these details in the Discussion section). The offensive team might not achieve developed coordination based on the tips of this role because of some factors such as only a single practice and individual skills.

3.2 Analysis procedures of playing related to the offensive role of intervention decision and adjustment

We obtained time-series position data of each participant in the two dimensions, x- and y-components, through the image processing of a bird's-eye view recording (Figure 2). This study compared the movements of offensive #3, required for the role of intervention decision and adjustment, as shown in pink in Figure 1, between the first and second halves using the dataset. The participants in the movies (20 fps, 1,280 px × 720 px) were tracked using ByteTrack (43) based on YOLOX. The experimenter manually corrected false detections and undetected positives using the labeling platform Labelbox (44) and prepared the dataset from real coordinate transformations. The average of the absolute measurement errors was 3.165 cm × 2.102 cm, similar to those (2.3 cm × 4.6 cm) reported in the previous study (18) that investigated passing coordination by a triad in soccer. Considering the bin interval of 200 cm in the histogram used in the analysis explained below, the errors in our study were unlikely to negatively influence the results. Meanwhile, in this practice, it is difficult to detect a ball using image processing technology due to the recording environment and technical problems; this is future work.

This study analyzed (1) the distance (cm) between the offensive participant required for the role of intervention decision and adjustment and each defensive player, as shown in black, orange, or yellow in Figure 1, and (2) the distance between the offensive key player and each other participant, as shown in red or blue in Figure 1. We evaluated spacing skill to play this role. Spacing skill, which maintains reasonable distances from other players, is generally crucial for both offensive and defensive coordination in basketball [e.g., (13, 14, 16, 45, 46)]. In the mini-game, if reasonable distances from the defensive players are maintained, the defensive pressure is reduced, which makes it easy to intervene moderately with the other offensive players according to the situations, as explained in Table 1. If those among the offensive players are maintained, it is easy to pass the ball to the relevant role, and the defense will break down. The characteristic is also represented by staying in place without interrupting the other offensive players, as shown in Table 1. This analysis calculated both indices of (1) and (2) for each time frame and made these histograms. Histogram analysis is a fundamental method to understand overall trends for investigating complex and dynamic coordination. It has been used in previous studies on group behavior [e.g., (4, 47, 48)] because of its advantage in intuitively capturing the characteristics of continuous values. Histograms were generated in each trial, and the normalized frequencies were averaged for the first and second halves. A t-test was conducted to compare them between the first and second halves in each bin. It is expected that significant differences in the frequencies for the indices of (1) and (2) are confirmed because of giving the tips of this role. This study conducted exploratory comparisons of reasonable distances between the participants because of analyzing complex and dynamic group behavior in the field of sports.

The previous studies (4, 48) conducted the t-test in each bin of a histogram to statistically compare frequencies between different conditions. However, we should keep in mind the type I error by the multiple comparisons. Therefore, this study calculated the effect sizes (Cohen's d) and powers (1-β) in all the bins for both indices of (1) and (2) to improve the validity of the statistical tests and carefully interpret these results. For calculating the powers, we set the significant level at 5% (α = 0.05), effect sizes, and sample size n (21 trials) in each half. We comprehensively evaluated the differences between the first and second halves in terms of the p-value, effect size, and power indices. The Bonferroni method was not applied to avoid this because the numbers of the bins in both indices of (1) and (2) were large and it was easy to cause the type II error. Our policy was to avoid this problem by calculating the three indices and investigating the significant differences between the first and second halves from different perspectives. The distances between the participants were analyzed using MATLAB R2021b. The statistical tests and the investigation of these validities were conducted using R 4.3.3 and G*Power 3.1, respectively. However, it is crucial to compare the playing related to the role of intervention decision and adjustment using applied statistical methods before and after giving the tips of this role; this is also future work.

3.3 Results of the distance with each defensive player

Figure 4 shows a time-series example of the distance (cm) between the offensive participant for the role of intervention decision and adjustment and each defensive player in a trial. Figure 4 also shows the histogram in each session. The upper and lower parts represent those in the first and second halves, respectively. The horizontal and vertical axes indicate the bin and average of the normalized frequencies across the seven trials of each session, respectively.

Figure 4

Figure 4. Overall characteristics of the distances (cm) between the offensive player required for the role of intervention decision and adjustment and each defensive participant. Its time-series example in a trial fluctuates up and down. Meanwhile, the upper and lower parts represent the histograms in the first and second halves, respectively. The horizontal and vertical axes indicate the bin and average of the normalized frequencies across the seven trials of each session, respectively.

In each bin, the average of normalized frequencies across the 21 trials in the first half from Sessions 1 to 3 was compared with that in the second one from Sessions 4 to 6. Notably, the t-tests confirmed the significant differences in the bins of 400 cm, 1,200 cm, and 1,400 cm, and these effect sizes ranged from medium to large while the powers were high [400 cm: t (20) = 2.739, p = 0.013, d = 0.894, 1-β = 0.973; 1,200 cm: t (20) = −2.269, p = 0.034, d = 0.732, 1-β = 0.891; 1,400 cm: t (20) = −2.581, p = 0.018, d = 0.668, 1-β = 0.829; Figures 5A–C). In the bin of a relatively small distance of 400 cm, the frequency in the second half was significantly lower than that in the first one. Conversely, in the bins of relatively large distances of 1,200 cm and 1,400 cm, the frequencies in the second half were significantly higher than those in the first one. In the other bins, no significant differences were confirmed (see these details in Supplementary Materials).

Figure 5

Figure 5. Averages of the normalized frequencies in the bins confirmed the significant differences between the first and second halves in the histograms of Figure 4. In the bin of a relatively small distance of 400 cm (A), the normalized frequency in the second half is significantly lower than that in the first one. Conversely, in the bins of relatively large distances of 1,200 cm (B) and 1,400 cm (C), the frequencies in the second half are significantly higher than those in the first one.

3.4 Results of the distance with each other offensive player

Figure 6 shows a time-series example of the distance (cm) between the offensive player required for the role of intervention decision and adjustment and each other participant in a trial. Figure 6 also shows the histogram in each session. The upper and lower parts represent those in the first and second halves, respectively. The horizontal and vertical axes indicate the bin and average of the normalized frequencies across the seven trials of each session, respectively.

Figure 6

Figure 6. Overall characteristics of the distances (cm) between the offensive player required for the role of intervention decision and adjustment and each other participant. Its time-series example in a trial fluctuates up and down. Meanwhile, the upper and lower parts represent the histograms in the first and second halves, respectively. The horizontal and vertical axes indicate the bin and average of the normalized frequencies across the seven trials of each session, respectively.

In each bin, the average of normalized frequencies across the 21 trials in the first half from Sessions 1 to 3 was compared with that in the second one from Sessions 4 to 6. Considering the results of the t-tests, effect sizes, and powers, the clear differences were not confirmed between the first and second halves. Although the effect sizes were medium and the powers were lower than those reported in the previous section, the t-tests indicated the significant trends of differences in the bins of 200 cm, 800 cm, and 1,600 cm [200 cm: t (20) = −2.002, p = 0.059, d = 0.517, 1-β = 0.616; 800 cm: t (20) = 1.856, p = 0.078, d = 0.628, 1-β = 0.782; 1,600 cm: t (20) = −1.745, p = 0.096, d = 0.572, 1-β = 0.703; Figures 7A–C). These showed a U-shape trend; in the bins of relatively small and large distances of 200 cm and 1,600 cm, respectively, the frequencies in the second half tended to be significantly higher than those in the first one. Conversely, in the middle bin of 800 cm, the frequency in the second half tended to be significantly lower than the first one. In the other bins, no significant differences were confirmed (see these details in Supplementary Materials).

Figure 7

Figure 7. Averages of the normalized frequencies in the bins confirmed the significant trends of differences between the first and second halves in the histograms of Figure 6. In the bins of relatively small and large distances of 200 cm (A) and 1,600 cm (C), respectively, the normalized frequencies in the second half tend to be significantly higher than those in the first one. Conversely, the frequency in the middle bin of 800 cm (B) in the second half tends to be significantly lower than that in the first one.

4 Discussion

This study developed the mini-game of 3-on-3 basketball, in which the offensive role of intervention decision and adjustment is crucial for winning. We introduced it to the practice of the female university team as a pilot study. The purpose of our study was to investigate the influence of this role on coordination. In this practice, the results confirmed that in the bins of the relatively large distance between the participant required for the relevant role and each defensive player, the frequencies after receiving the tips of the crucial role were significantly higher than before (Figures 5B,C). Conversely, in the bin of the relatively small distance, the frequency after receiving these tips was significantly lower than before (Figure 5A). Furthermore, the winning rate on the offensive team improved temporarily (Figure 3); however, the effects of receiving the tips were not maintained. Therefore, these results partially supported the hypothesis regarding the effects of giving the tips.

After receiving the tips, the participant required for the offensive role of intervention decision and adjustment maintained a reasonable distance from each defensive participant. This suggests that the spacing skill might create favorable situations for coordination, such as moving to the corner area and aiming to take a three-point shot, as explained in Table 1 (Figure 8). Although the effect sizes and powers were lower for the distance with the other offensive players compared with those with the defensive players, notable trends were observed. In the bins of relatively small and large distances, the frequencies after receiving the tips of this role tended to be significantly higher than before (Figures 7A,C). In the bin of the middle distance, the converse result was obtained (Figure 7B). Such a U-shape trend might indicate that the participant required for the relevant role also stayed in place without interrupting the other offensive participants, as explained in Table 1. However, these discussions should be investigated in more detail. The previous studies in basketball suggested that offensive interactions, involving ball movement to an opposite side and maintaining a reasonable distance with a defensive player, create a gap with the opponent (16, 45). These lead to offensive shooting opportunities in open space. In this practice, running to the corner area can serve as a starting point for moving the ball to the opposite side. This enabled the participant required for the crucial role of taking a three-point shot in an open space. Therefore, such coordination increases the countermeasures for the defensive team, making it difficult to anticipate the next playing. Spacing skill, including a reasonable distance from the goal, is related to a high success rate of offensive shooting and defense in basketball. Expanding space within one's own team and limiting it for the opponent team to develop successful opportunities. These variables are key to predicting team performance using statistical and machine learning models [e.g., (13, 14, 16, 45, 46)]. It is also essential to create favorable situations in soccer other than basketball and influence strategies [e.g., (49, 50)]. The previous study (49) confirmed that the experimenter manipulated space to play in a soccer game and its limitations influence strategic behaviors. Another study (50) quantified and visualized the occupied space of each soccer player using applied geometric analysis with Voronoi diagrams. These results suggest that dynamical adjustment of defensive space according to situations limits opponent passing options. Although the types of team sports and methods used in these studies differed from those used in this study, the findings are similar to ours. Meanwhile, many previous studies mentioned above analyzed professional teams. Hence, our results from the female university team, which was affiliated with the third division league in Japan and did not have a high skill level, are valuable. Therefore, this study provides new insights into spacing skill related to the role of intervention decision and adjustment.

Figure 8

Figure 8. Typical example of offensive coordination after receiving the tips about the role of intervention decision and adjustment, observed in this practice. #1 runs toward the area under the basket goal (“paint area”) and receives a pass from #2. Subsequently, #3 required for this role moves to the corner area while maintaining a reasonable distance from each defensive player, receives a pass from #1, and aims to take a three-point shot.

This study may also provide the results that satisfy the usefulness and ecological validity of the role of intervention decision and adjustment. Ecological validity is related to the problem that the experimental environment and observed actions significantly differ from real-world activities (51). However, after receiving the tips for this role, high offensive team performance could not be maintained in the final session (Session 6). This was caused by the defensive team establishing an effective countermeasure. After this practice, the experimenter asked the defensive team about it; the participants reported that defensive #3 marking the participant in the relevant role pretended to help the other players while actually kept marking. Thus, it was easy to establish the defensive countermeasure because the offensive pattern, as shown in Figure 8, was standardized. The offensive team had to flexibly handle the defensive one through bargaining based on the tips. However, the participants could not achieve developed coordination because of some factors, such as only a single practice and their individual skills corresponding to the third division league in Japan; they would need more training. Flexible coordination according to dynamic situations is required in various activities regardless of team sports (31, 52–54). Furthermore, both top-down and bottom-up information processing are crucial for coordination. The shared mental model of knowledge structures including strategies leads to efficient group interactions because the members can explain and anticipate actions from each other. A group also needs to fine-tune to a dynamically changing environment and others' unanticipated actions (55–57). Defensive coordination in 5-on-5 basketball games for the top-level university team in Japan shows that the structures of role-sharing switch according to emergent situations (1). In this practice, the offensive team established a shared mental model based on the tips of the crucial role. Although such top-down processing worked, developed bottom-up one might not be established in this team.

Future studies should mainly investigate (1) the number of teams participating in this practice, (2) the direct evaluation of spacing skill, and (3) these applications in other team sports. Regarding (1), only one female university team participated in this practice. The mini-games must be conducted for other teams with similar skill level. Furthermore, it is crucial to conduct the games for expert teams. The game requires basic coordinated behavior. Hence, if expert teams play it, they can express key strategies without the tips from a coach. Next as for (2), this study highlights the importance of spacing skill related to the role of intervention decision and adjustment. However, it is necessary to evaluate it directly. Similar to the previous study (50), applied geometric analysis with Voronoi diagrams enables the quantification and visualization of the occupied space of each player and the investigation of systematic offensive coordination with the player in this role. Additionally, analysis considering contexts, such as passing, dribbling, and shooting, is required [e.g., (13–16)]. These approaches develop deep discussions; it should be kept in mind that the results of this study indicate the overall trends of playing related to the particular role based on the experimental findings in cognitive science (21). In relation to (3), as mentioned above, spacing skill is also required in 5-on-5 basketball and soccer. Hence, if we develop a similar mini-game for other team sports, in which the relevant role is key, practical applications might be conducted. The academic and social impacts are significant because we provide an example of how to bridge the gap between controlled experiments and real-world applications. Therefore, our findings may contribute to practice design for the acquisition of spacing skill related to the relevant role and off-ball movements. Our work may offer implications for further developing 3-on-3 basketball itself. Similar to the experiment (49), a practice may be developed where the play range is limited, encouraging attention to space. For example, a coach can manipulate the offensive play range by changing the positions of defensive opponents. If a defensive player approaches to help another player, an offensive player in the role of intervention decision and adjustment role runs to a corner area and takes a three-point shot. Meanwhile, if a defensive opponent does not approach, the offensive player does not pass to this role but keeps dribbling to take a shot.

5 Conclusion

We applied the experimental findings in the cognitive science study of the crucial role in coordinated behavior of a triad through role-sharing to the field of sports as a pilot study. This study investigated the influence of the offensive role of intervention decision and adjustment on coordination in 3-on-3 basketball to discuss the usefulness of this role. The mini-game was developed, in which the relevant role is key, and introduced it to the practice of the female university team. The results showed that in the bins of the relatively large distance between the participant required for this role and each defensive player, the frequencies after receiving the tips of the mentioned role were significantly higher; the winning rate on the offensive team improved temporarily; however, the effects were not maintained. This suggests that spacing skill, which maintains a reasonable distance from each defensive player, emerged. Our study may provide the findings that satisfy the usefulness and ecological validity. These may bridge the gap between controlled experiments and real-world applications and contribute to practice design for the acquisition of such spacing skill and off-ball movements. In future work, it is important to examine the interactive structure to discuss the offensive coordination process.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Ethics statement

The studies involving humans were approved by the ethics and safety committee of Shizuoka University and Tokoha University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions

JI: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Software, Supervision, Visualization, Writing – original draft, Writing – review & editing. MY: Conceptualization, Formal analysis, Investigation, Methodology, Resources, Writing – review & editing. KF: Conceptualization, Formal analysis, Methodology, Writing – review & editing.

Funding

The authors declare that financial support was received for the research and/or publication of this article. This study was supported by JSPS KAKENHI Grant Numbers 24K20562, AY 2023 (Special Application) Research Grants of Amano Institute of Technology, and AY 2023 Research Grants of Establishment of Interdisciplinary Networks for Development of Innovative Areas in Shizuoka University.

Acknowledgments

We would like to thank Naruhisa Takamura, who is a coach at Beltex Shizuoka Academy, and Masanari Ichikawa of Shizuoka University for their helpful cooperation with the practice. We would also like to express our gratitude to Genki Ichinose and Yugo Takeuchi of Shizuoka University for making meaningful comments. This paper is published on the preprint server bioRxiv (58).

Conflict of interest

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.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fspor.2025.1513982/full#supplementary-material

References

1. Fujii K, Yokoyama K, Koyama T, Rikukawa A, Yamada H, Yamamoto Y. Resilient help to switch and overlap hierarchical subsystems in a small human group. Sci Rep. (2016) 6:23911. doi: 10.1038/srep23911

PubMed Abstract | Crossref Full Text | Google Scholar

2. Takagi A, Hirashima M, Nozaki D, Burdet E. Individuals physically interacting in a group rapidly coordinate their movement by estimating the collective goal. eLife. (2019) 8:e41328. doi: 10.7554/eLife.41328.001

PubMed Abstract | Crossref Full Text | Google Scholar

3. Woolley AW, Chabris CF, Pentland A, Hashmi N, Malone TW. Evidence for a collective intelligence factor in the performance of human groups. Science. (2010) 330(6004):686–8. doi: 10.1126/science.1193147

PubMed Abstract | Crossref Full Text | Google Scholar

4. Yokoyama K, Shima H, Fujii K, Tabuchi N, Yamamoto Y. Social forces for team coordination in ball possession game. Phys Rev E. (2018) 97(2):022410. doi: 10.1103/PhysRevE.97.022410

PubMed Abstract | Crossref Full Text | Google Scholar

5. Knoblich G, Butterfill S, Sebanz N. Psychological research on joint action: theory and data. In: Ross BH, editor. Psychology of Learning and Motivation. Cambridge: Elsevier Academic Press (2011). p. 59–101.

Google Scholar

6. Hollan J, Hutchins E, Kirsh D. Distributed cognition: toward a new foundation for human-computer interaction research. ACM Trans Comput Hum Interact. (2000) 7(2):174–96. doi: 10.1145/353485.353487

Crossref Full Text | Google Scholar

7. Hutchins E. Cognition in the Wild. Cambridge: MIT press (1995).

Google Scholar

8. Hayashi Y. The power of a “Maverick” in collaborative problem solving: an experimental investigation of individual perspective-taking within a group. Cognit Sci. (2018) 42(S1):69–104. doi: 10.1111/cogs.12587

Crossref Full Text | Google Scholar

9. Hayashi Y, Miwa K, Morita J. A laboratory study on distributed problem solving by taking different viewpoints. Proc Cog Sci. (2006). p. 333–8.

Google Scholar

10. Shirouzu H, Miyake N, Masukawa H. Cognitively active externalization for situated reflection. Cognit Sci. (2002) 26(4):469–501. doi: 10.1207/s15516709cog2604_3

Crossref Full Text | Google Scholar

11. Sun C, Shute VJ, Stewart A, Yonehiro J, Duran N, D’Mello S. Towards a generalized competency model of collaborative problem solving. Comput Educ. (2020) 143:103672. doi: 10.1016/j.compedu.2019.103672

Crossref Full Text | Google Scholar

12. Zhang Z, Bekker T, Markopoulos P, Skovbjerg HM. Supporting and understanding students’ collaborative reflection-in-action during design-based learning. Int J Technol Des Educ. (2024) 34:307–43. doi: 10.1007/s10798-023-09814-0

Crossref Full Text | Google Scholar

13. Kono R, Fujii K. Mathematical models for off-ball scoring prediction in basketball. arXiv [Preprint]. arXiv:2406.08749 (2024). http://arxiv.org/abs/2406.08749 (Accessed September 04, 2024).

Google Scholar

14. Lamas L, Santana F, Heiner M, Ugrinowitsch C, Fellingham G. Modeling the offensive-defensive interaction and resulting outcomes in basketball. PLoS One. (2015) 10(12):e0144435. doi: 10.1371/journal.pone.0144435

PubMed Abstract | Crossref Full Text | Google Scholar

15. Supola B, Hoch T, Baca A. Modeling the extra pass in basketball–an assessment of one of the most crucial skills for creating great ball movement. Int J Comput Sci Sport. (2023) 22(1):13–29. doi: 10.2478/ijcss-2023-0002

Crossref Full Text | Google Scholar

16. Wu Y, Deng D, Xie X, He M, Xu J, Zhang H, et al. OBTracker: visual analytics of off-ball movements in basketball. IEEE Trans Visual Comput Graphics. (2022) 29(1):929–39. doi: 10.1109/TVCG.2022.3209373

PubMed Abstract | Crossref Full Text | Google Scholar

17. Ichikawa J, Fujii K. Proposal of a research approach for discussion of a dynamic coordination mechanism: investigation of anticipating others’ behaviors and adaptation through quantitative analysis of group behavior. Cognit Stud Bull Jpn Cogn Sci Soc. (2020) 27(3):377–85. doi: 10.11225/cs.2020.026

Crossref Full Text | Google Scholar

18. Yokoyama K, Yamamoto Y. Three people can synchronize as coupled oscillators during sports activities. PLoS Comput Biol. (2011) 7(10):e1002181. doi: 10.1371/journal.pcbi.1002181

PubMed Abstract | Crossref Full Text | Google Scholar

19. Braun DA, Ortega PA, Wolpert DM. Nash equilibria in multi-agent motor interactions. PLoS Comput Biol. (2009) 5(8):e1000468. doi: 10.1371/journal.pcbi.1000468

PubMed Abstract | Crossref Full Text | Google Scholar

20. Chackochan VT, Sanguineti V. Incomplete information about the partner affects the development of collaborative strategies in joint action. PLoS Comput Biol. (2019) 15(12):e1006385. doi: 10.1371/journal.pcbi.1006385

PubMed Abstract | Crossref Full Text | Google Scholar

21. Ichikawa J, Fujii K. Analysis of group behavior based on sharing heterogeneous roles in a triad using a coordinated drawing task. Front Psychol. (2022) 13:890205. doi: 10.3389/fpsyg.2022.890205

PubMed Abstract | Crossref Full Text | Google Scholar

22. Fujii K, Kawasaki T, Inaba Y, Kawahara Y. Prediction and classification in equation-free collective motion dynamics. PLoS Comput Biol. (2018) 14(11). doi: 10.1371/journal.pcbi.1006545

PubMed Abstract | Crossref Full Text | Google Scholar

23. Fujii K, Takeuchi K, Kuribayashi A, Takeishi N, Kawahara Y, Takeda K. Estimating counterfactual treatment outcomes over time in complex multiagent scenarios. IEEE Trans Neural Networks Learn Syst. (2024) 36(2):2103–17. doi: 10.1109/TNNLS.2024.3361166

PubMed Abstract | Crossref Full Text | Google Scholar

24. Nakahara H, Tsutsui K, Takeda K, Fujii K. Action valuation of on-and off-ball soccer players based on multi-agent deep reinforcement learning. IEEE Access. (2023) 11:131237–44. doi: 10.1109/ACCESS.2023.3336425

Crossref Full Text | Google Scholar

25. Buldú JM, Busquets J, Echegoyen I, Seirul Lo F. Defining a historic football team: using network science to analyze Guardiola’s F.C. Barcelona. Sci Rep. (2019) 9:13602. doi: 10.1038/s41598-019-49969-2

Crossref Full Text | Google Scholar

26. Clemente FM, Sarmento H, Aquino R. Player position relationships with centrality in the passing network of world cup soccer teams: win/loss match comparisons. Chaos Solitons Fractals. (2020) 133:109625. doi: 10.1016/j.chaos.2020.109625

Crossref Full Text | Google Scholar

27. Ichinose G, Tsuchiya T, Watanabe S. Robustness of football passing networks against continuous node and link removals. Chaos Solitons Fractals. (2021) 147:110973. doi: 10.1016/j.chaos.2021.110973

Crossref Full Text | Google Scholar

28. Gazda SK, Connor RC, Edgar RK, Cox F. A division of labour with role specialization in group-hunting bottlenose dolphins (Tursiops truncatus) off cedar key. Florida. Proc R Soc B Biol Sci. (2005) 272(1559):135–40. doi: 10.1098/rspb.2004.2937

Crossref Full Text | Google Scholar

29. Stander PE. Behavioral ecology and sociobiology cooperative hunting in lions: the role of the individual. Behav Ecol Sociobiol. (1992) 29:445–54. doi: 10.1007/BF00170175

Crossref Full Text | Google Scholar

30. Tsutsui K, Tanaka R, Takeda K, Fujii K. Collaborative hunting in artificial agents with deep reinforcement learning. eLife. (2024) 13:e85694. doi: 10.7554/eLife.85694

PubMed Abstract | Crossref Full Text | Google Scholar

31. Bowers C, Kreutzer C, Cannon-Bowers J, Lamb J. Team resilience as a second-order emergent state: a theoretical model and research directions. Front Psychol. (2017) 8:1360. doi: 10.3389/fpsyg.2017.01360

PubMed Abstract | Crossref Full Text | Google Scholar

32. Burke CS, Stagl KC, Salas E, Pierce L, Kendall D. Understanding team adaptation: a conceptual analysis and model. J Appl Psychol. (2006) 91(6):1189–207. doi: 10.1037/0021-9010.91.6.1189

PubMed Abstract | Crossref Full Text | Google Scholar

33. Johnston JH, Phillips HL, Milham LM, Riddle DL, Townsend LN, DeCostanza AH, et al. A team training field research study: extending a theory of team development. Front Psychol. (2019) 10:1480. doi: 10.3389/fpsyg.2019.01480

PubMed Abstract | Crossref Full Text | Google Scholar

34. Palmer C. A theory of risk and resilience factors in military families. Mil Psychol. (2008) 20(3):205–17. doi: 10.1080/08995600802118858

Crossref Full Text | Google Scholar

35. Richardson MJ, Marsh KL, Isenhower RW, Goodman JRL, Schmidt RC. Rocking together: dynamics of intentional and unintentional interpersonal coordination. Hum Mov Sci. (2007) 26(6):867–91. doi: 10.1016/j.humov.2007.07.002

PubMed Abstract | Crossref Full Text | Google Scholar

36. Schmidt RC, Carello C, Turvey MT. Phase transitions and critical fluctuations in the visual coordination of rhythmic movements between people. J Exp Psychol Hum Percept Perform. (1990) 16(2):227–47. doi: 10.1037/0096-1523.16.2.227

Crossref Full Text | Google Scholar

37. Shimizu D, Okada T. Synchronization and coordination of art performances in highly competitive contexts: battle scenes of expert breakdancers. Front Psychol. (2021) 12:635534. doi: 10.3389/fpsyg.2021.635534

PubMed Abstract | Crossref Full Text | Google Scholar

38. Walton AE, Washburn A, Langland-Hassan P, Chemero A, Kloos H, Richardson MJ. Creating time: social collaboration in music improvisation. Top Cognit Sci. (2018) 10(1):95–119. doi: 10.1111/tops.12306

PubMed Abstract | Crossref Full Text | Google Scholar

39. Hojo M, Fujii K, Inaba Y, Motoyasu Y, Kawahara Y. Automatically recognizing strategic cooperative behaviors in various situations of a team sport. PLoS One. (2018) 13(12):e0209247. doi: 10.1371/journal.pone.0209247

PubMed Abstract | Crossref Full Text | Google Scholar

40. Spearman W. Beyond expected goals. Proc. 12th MIT Sloan Sports Anal. Conf. (2018). p. 1–17

Google Scholar

41. Japan Basketball Association. 2022 Official 3×3 Basketball Rules. (2022). Available at: http://www.japanbasketball.jp/files/referee/rule/2022_3×3rule_20240401.pdf (Accessed March 17, 2025).

Google Scholar

42. Japan Basketball Association. 2022 Official Basketball Rules. (2023). Available at: http://www.japanbasketball.jp/files/referee/rule/2023rule.pdf (Accessed January 25, 2023).

Google Scholar

43. Zhang Y, Sun P, Jiang Y, Yu D, Weng F, Yuan Z, et al. ByteTrack: Multi-object tracking by associating every detection box. arXiv [Preprint]. arXiv:2110.06864 (2022). Available at: https://arxiv.org/abs/2110.06864 (Accessed June 26, 2024).

Google Scholar

44. Labelbox. Labelbox The Data Factory for Next Gen AI (2018). Available at: https://labelbox.com (Accessed April 26, 2024).

Google Scholar

45. Esteves PT, Silva P, Vilar L, Travassos B, Duarte R, Arede J, et al. Space occupation near the basket shapes collective behaviours in youth basketball. J Sports Sci. (2015) 34(16):1557–63. doi: 10.1080/02640414.2015.1122825

PubMed Abstract | Crossref Full Text | Google Scholar

46. Franks A, Miller A, Bornn L, Goldsberry K. Characterizing the spatial structure of defensive skill in professional basketball. Ann Appl Stat. (2015) 9(1):94–121. doi: 10.1214/14-AOAS799

Crossref Full Text | Google Scholar

47. Cavagna A, Cimarelli A, Giardina I, Parisi G, Santagati R, Stefanini F, et al. Scale-free correlations in starling flocks. Proc Natl Acad Sci U S A. (2010) 107(26):11865–70. doi: 10.1073/pnas.1005766107

PubMed Abstract | Crossref Full Text | Google Scholar

48. Ichikawa J, Fujii K, Nagai T, Omori T, Oka N. Quantitative analysis of spontaneous sociality in children’s group behavior during nursery activity. PLoS One. (2021) 16(2):e0246041. doi: 10.1371/journal.pone.0246041

PubMed Abstract | Crossref Full Text | Google Scholar

49. Gonçalves B, Esteves P, Folgado H, Ric A, Torrents C, Sampaio J. Effects of pitch area-restrictions on tactical behavior, physical, and physiological performances in soccer large-sided games. J Strength Cond Res. (2017) 31(9):2398–408. doi: 10.1519/JSC.0000000000001700

PubMed Abstract | Crossref Full Text | Google Scholar

50. Taki T, Hasegawa J. Visualization of dominant region in team games and its application to teamwork analysis. Proc Comput Graph Int. (2000). p. 227–35. doi: 10.1109/CGI.2000.852338

Crossref Full Text | Google Scholar

51. Neisser U. Cognition and Reality: Principles and Implications of Cognitive Psychology. San Francisco: W H Freeman/Times Books Henry Holt & Co (1976).

Google Scholar

52. Amon MJ, Vrzakova H, D’Mello SK. Beyond dyadic coordination: multimodal behavioral irregularity in triads predicts facets of collaborative problem solving. Cognit Sci. (2019) 43(10):e12787. doi: 10.1111/cogs.12787

PubMed Abstract | Crossref Full Text | Google Scholar

53. Gorman JC, Cooke NJ, Amazeen PG, Fouse S. Measuring patterns in team interaction sequences using a discrete recurrence approach. Hum Factors. (2011) 54(4):503–17. doi: 10.1177/0018720811426140

Crossref Full Text | Google Scholar

54. Kijima A, Shima H, Okumura M, Yamamoto Y, Richardson MJ. Effects of agent-environment symmetry on the coordination dynamics of triadic jumping. Front Psychol. (2017) 8:3. doi: 10.3389/fpsyg.2017.00003

PubMed Abstract | Crossref Full Text | Google Scholar

55. Cooke NJ, Gorman JC, Myers CW, Duran JL. Interactive team cognition. Cognit Sci. (2013) 37(2):255–85. doi: 10.1111/cogs.12009

Crossref Full Text | Google Scholar

56. Gorman JC. Team coordination and dynamics: two central issues. Curr Direct Psychol Sci. (2014) 23(5):355–60. doi: 10.1177/0963721414545215

Crossref Full Text | Google Scholar

57. Steiner S, Macquet AC, Seiler R. An integrative perspective on interpersonal coordination in interactive team sports. Front Psychol. (2017) 8:1440. doi: 10.3389/fpsyg.2017.01440

PubMed Abstract | Crossref Full Text | Google Scholar

58. Ichikawa J, Yamada M, Fujii K. Analyzing coordinated group behavior through role-sharing: A pilot study in female 3-on-3 basketball with practical application. bioRxiv [Preprint] (2024). Available at: https://www.biorxiv.org/content/10.1101/2024.09.16.612561v3 (Accessed September 16, 2024).

Google Scholar