NBA

NBA analytics are in development

The same percentile framework that powers the NCAA and international boards, applied to the NBA.

Not live yet — this page will fill up as the data pipeline is built.

Gaussian Hoops currently covers NCAA Division I and fifteen international professional leagues. The NBA is the missing link: it is where the prospects on every other board end up, and it is the benchmark the draft model is trained against.

What is planned

  • 1.NBA as its own tier. Percentiles stay strictly separated — NBA players are ranked against NBA players, never mixed with college or international pools. Comparing a rookie's rebound rate to a mid-major centre's would be meaningless.
  • 2.Weekly updates during the season, so the boards reflect the current campaign rather than a frozen snapshot.
  • 3.Cross-league comparisons. Which NBA players a prospect statistically resembles — the natural extension of the Similar Players tool.
  • 4.Draft model validation. The star probabilities on the draft boards are trained on what actually happened in the NBA. More NBA data means a sharper model.

In the meantime

The draft class pages already carry NBA outcomes for every pick from 2021 onward — games played, peak minutes, peak production — set against what the model predicted before the draft. That is the closest thing to NBA data on the site today.