German 3x3 Basketball Player Assessments and Coach Selection Decisions

Motor performance, anthropometric, and subjective coach assessment data for 46 German youth 3x3 basketball players. The dataset supports a methodological comparison of logistic regression and fast-and-frugal tree models for predicting player selection. Nora Cermak authored the study, with data last updated in March 2026.

Use Cases

• Predict player selection using logistic regression on features like motor performance tests and subjective coach assessments.
• Model coach decision heuristics by constructing a fast-and-frugal tree from cues like technical, tactical, and psychological parameter ratings.
• Analyze interrater reliability between coaches using intraclass correlation coefficients calculated from subjective assessment data.
• Compare the relevance of talent aspects, such as current performance, between national and regional coaches using graphical analyses and descriptive statistics.

Strengths

• Data includes 46 players with a balanced gender split (23 female).
• Contains multidimensional data: motor tests, anthropometrics, and subjective coach ratings across technical, tactical, and psychological parameters.
• Study compares two distinct statistical modeling approaches (logistic regression and FFTree) on the same data.

Limitations

• Small sample size of 46 players limits statistical power and generalizability.
• Interrater reliability between coaches was poor to moderate (ICC 0.38-0.51), indicating high subjectivity.
• Geographic scope is limited to the German national selection system.

Provenance

Source

Nora Cermak via figshare.

Collection Method

Data collected during the final stage of the German national 3x3 basketball selection system, including performance tests and coach assessments.

Time Range

Data collection period not specified, but related to the 2026 publication.

Freshness

Data was last updated in March 2026.

Geography

Germany.