Author(s): CHEN, C., LIU, Y.T., Institution: NATIONAL TAIWAN NORMAL UNIVERSITY, Country: TAIWAN, Abstract-ID: 1928

INTRODUCTION:
In basketball, the ball handler’s decision-making (DM) has a substantial influence on scoring and winning. Previous studies have suggested that defensive performance may serve as an indicator of the ball handler’s DM quality. However, within a single possession (PO), the ball handler takes more than one action, and the number of these actions seems to relate to game results. Therefore, this study examined team performance in 3×3 basketball by evaluating the ball handler’s cumulative DM performance within each PO.
METHODS:
40 games from the FIBA women’s 3×3 U23 World Cup were analyzed. Simi Scout was used to record ball-handler actions, including dribbling, drives, passes, and shots. Defensive outcomes were quantified to evaluate ball-handler DM, with turnover as −1, effective defensive positioning as 0, defensive failure as 1, and opponent fouls as 0.5. Shot outcomes were also recorded. The total number of actions in a PO was used to measure PO length, and POs were further classified as short (1–2), medium (3–4), or long (5–10). The accumulated defensive scores were then divided by PO length to obtain the PO decision score. Inter- and intra-rater reliability of the recorded events were assessed using Cohen’s kappa. Shot rate, goal rate per shot, goal rate per PO, and PO decision score were analyzed using a 2 (win, loss) by 3 (short, medium, long) repeated-measures ANOVA in Jamovi.
RESULTS:
Cohen’s kappa ranged from .727 to .995. A significant main effect of outcome was observed across all dependent variables, with winners outperforming losers, ps<.001. For PO decision scores, shot rate, and goal rate per PO, PO length was also significant, ps<.01, with short POs showing the lowest performance. Medium POs outperformed long POs for PO decision scores, whereas the shot rate showed the opposite pattern, and the goal rate per PO showed no difference between the two. No interaction effect was observed.
CONCLUSION:
This study examined how the quality of ball-handler DM influences game performance in 3×3 basketball, using PO as the unit of analysis. The results showed that winning teams achieved significantly higher PO decision scores, shot rates, and goal rates than losing teams, indicating that ball-handler DM is closely related to game outcomes. Regarding PO length, POs of 3–4 actions yielded the highest decision scores and outperformed POs with both fewer and more actions. This finding suggests that effective ball-handler DM facilitates offensive success under a moderate-length-of-action organization. In contrast, poorer decisions tend to result in prematurely terminated short POs. POs with five or more actions had a higher shot rate than those with 3–4 actions, while achieving a comparable goal rate. This may indicate that DM during the final action before the end of a PO influences both shot attempts and scoring outcomes. Future work should examine the composition of ball-handler action sequences and their impact on game performance.