THE BREAKOUT BOARD
Which young players take a leap in 2026-27 — ranked by projected Net Impact gain. A least-squares model on five seasons of trajectories (rising usage, minutes, and efficiency, weighted by age), anchored to each player's true current Net Impact.
Projecting one-year leaps is noisy by nature, so the value is in the ranking — and the bands are honest (±3.5 Net Impact). Similarity, not destiny. As of 2026-07-31.
How the projections work
We first learn a box-score → Net Impact map on the current season (R²≈0.95) so older seasons that predate Net Impact can be scored on the same scale. For each player we measure year-over-year slopes in usage, minutes, efficiency, and that proxy Net Impact, plus career stage.
A least-squares model fit on every historical season transition predicts next-year change. We hold out the most recent transition to report a real hit-rate (above) and the residual spread becomes the confidence band. Young players (≤6 seasons) with a real role are surfaced; we rank by projected gain.