Metric · NBA

Offensive Impact

v2.0.0

Publish the offensive component of the same role-relative, minute-shrunk descriptive index used by UB Impact.

O_{r a w} = 0.34 z (P S A_{36}) + 0.20 z (U S G %) + 0.18 z (A S T %) + 0.12 z (A S T_{36}) + 0.08 z (O R E B_{36}) - 0.08 z (T O V %)

Offensive inputs are standardized within the declared role bucket.

What it measures

Scoring value, creation burden, and offensive box-score activity are standardized within role bucket before exposure shrinkage.

How to read it

Higher values indicate stronger descriptive offensive impact.

Why it exists

Publishing the halves makes the total auditable. The same eligible population, role-bucket standardization, minute shrinkage, and 0-100 scale are used for all three outputs. The halves are diagnostic components; the total is the only governed blend.

Limits

This component is diagnostic; the published total requires both components and uses the governed 0.60/0.40 blend.
Component weights are chosen, not fitted.

Calculation, evidence, and implementation

Calculation

O = c l am p (50 + 10 \frac{M I N}{M I N + 1200} O_{r a w}, 0, 100)

The component uses the same minute exposure shrinkage and display scale as UB Impact.

Inputs

Psa 36, Usg Pct, Ast Pct, Ast 36, Oreb 36, Tov Pct, Min, Gp

Methodology

Role-bucket z-score aggregation with minute shrinkage; component output from the canonical NBA player-season ETL.

Normalization

0-100 index with 50 as the eligible-population center.

Uncertainty

No confidence interval is published; eligibility and exposure are the available reliability signals.

Evaluation

Directionality and stability checks; not a causal or predictive estimate.

Sources

Pipelines/NBA Ingest/Stats.Py

Implementation

Registry identifier: NBA_IMPACT_OFFENSE. Version 2.0.0; status drafted.

Metric distribution →General methodology →