Fig 1.
Diagrams of 3-on-3 basketball mini-game in the field experiment to examine offensive coordination.
These illustrations are redesigned in the previous study [12] based on CC-BY. The original explanations are referred from the previous study. Circles and squares represent offensive and defensive players, respectively. The initial position of each offensive player is instructed in order to maintain certain distances among the players by the main experimenter. The defensive players (#1–#3) match up against the offensive players (#1–#3). The defensive players’ initial positions are self-selected. Just before each game, the experimenter instructs the two-man game position randomly on the right or left side of the goal. The ball handler is always offensive #1, and each game starts without direct passing to offensive #3. To win the mini-game within a 15 s time limit, offensive #3 should implement the key role of intervention decision and adjustment. This player needs to intervene with the teammates and adjust the two-man game to create favorable situations, such as (1) “hand-off,” (2) “pick and roll,” and (3) running to a corner area to attempt a three-point shot. Furthermore, it is crucial for offensive #3 (4) to stay in place without interrupting the teammates to ensure their play space. Solid arrows represent the movements of the off-ball player, zigzag arrows present those of the on-ball player, and a dotted arrow indicates the pass trajectory.
Fig 2.
Environment in the field experiment.
The upper part is redesigned from the figure in the previous study [12] based on CC-BY. The court area in the university gymnasium conforms to the official size of a 3-on-3 basketball court [45], except for the sideline. The video camera is set on the stage and covered the entire area and all the players from a bird’s-eye view. They are tracked using the image processing algorithm ByteTrack [48]. Written informed consent was obtained from all the players. According to this consent, the image is shown while blurring to avoid including individual identifiers.
Table 1.
Profiles of the offensive and defensive teams in the high and low expertise conditions.
Fig 3.
A typical example of judgment in the mini-game.
In Trial 1 of Session 2, offensive #3 wearing a pink bib took a three-point shot, and the two evaluators on the first step judged a win for the offensive team. A dotted arrow indicates the pass trajectory. Written informed consent was obtained from all the players. According to this consent, the image is shown while blurring to avoid including individual identifiers.
Fig 4.
Analysis procedures of entropy representing the difficulty in anticipating offensive movements for the defensive team.
A higher entropy indicates that the movement is more complex with higher uncertainty, making it difficult to anticipate.
Fig 5.
Analysis overview of the linear mixed model.
The dependent variable is entropy, which represents the difficulty in anticipating offensive movements for the defensive team, or the coefficient of variation in the distance (cm) between offensive #3 in the key role of intervention decision and adjustment and each other player. The main independent variable is the categorical condition. The model descriptions are according to the coding of the R packages.
Table 2.
Offensive team performance in the high expertise, low expertise before the tips, and low expertise after the tips conditions.
Fig 6.
Entropy representing the difficulty in anticipating offensive movements for the defensive team.
The horizontal and vertical axes present the role and entropy, respectively. For each role, the bars show the averages across the 21 trials for the four conditions. The gray error bars indicate the standard deviations, and the black error bars indicate 95% confidence intervals.
Table 3.
Likelihood ratio tests and AIC values using the linear mixed model for estimating the relationship between the condition and entropy. The entropy represents the difficulty in anticipating offensive movements for the defensive team. The values in the optimal model are shown in bold. The detailed specifications of each model are shown in Fig 5.
Table 4.
Optimal model estimating the relationship between the condition and entropy which represents the difficulty in anticipating offensive movements for the defensive team. The values of the fixed effects whose p-values are below the significance level (α = 0.05) are shown in bold. The low expertise before the tips condition and role O1 are set as the baseline categories.
Table 5.
Pairwise comparisons of the entropy in the optimal model which represents the difficulty in anticipating offensive movements for the defensive team between the conditions for each role. The values of the fixed effects whose -values are below the significance level (α = 0.05) are shown in bold.
Fig 7.
Coefficient of variation (CV) representing the variability in the distance (cm) between offensive #3 in the key role of intervention decision and adjustment and each other player.
The horizontal and vertical axes represent the pair with offensive #3 and CV, respectively. For each pair, the bars show the averages across the 21 trials for the three expertise conditions. The gray error bars indicate the standard deviations, and the black error bars indicate 95% confidence intervals.
Table 6.
Likelihood ratio tests and AIC values using the linear mixed model for estimating the relationship between the condition and the coefficient of variation (CV). The CV represents the variability in the distance (cm) between offensive #3 in the key role of intervention decision and adjustment and each other player. The values in the optimal model are shown in bold. The detailed specifications of each model are shown in Fig 5.
Table 7.
Optimal model estimating the relationship between the condition and the coefficient of variation (CV). The CV represents the variability in the distance (cm) between offensive #3 in the key role of intervention decision and adjustment and each other player. The values of the fixed effects whose -values are below the significance level (α = 0.05) are shown in bold. The low expertise before the tips condition and the pair of offensive #3 and offensive #1 are set as the baseline categories.
Fig 8.
Example of team strategies observed in the high expertise condition.
Trial 2 of Session 1 lasted about 8 s and resulted in a win for the offensive team. A solid arrow represents the movements of the off-ball player, zigzag arrows present those of the on-ball player, and a dotted arrow indicates the pass trajectory.