Models / LightGBM + quantile heads

Pace Normalization (LTOE)

Lap Time Over Expected — the normalized pace delta, per lap.

LightGBM + quantile headsanalytics.strategy

How it works

A LightGBM model predicts the expected lap time from fuel-corrected pace and stint context; the residual (actual − expected) is the LTOE signal — negative means genuinely faster. It feeds the pace-mode classification (PUSH / HOLD / MANAGE / WARM_UP / DELTA) and the true-pace ranking.

Inputs

  • fuel-corrected lap time
  • stint context
  • compound & tyre age
  • track evolution

Output

drivers[].ltoe.ltoeSecSeconds vs expected; negative = faster.
drivers[].ltoe.confidenceScale1.0 once 3+ races of the season's data exist.

Calibration

Season-early damper (confidenceScale 0.75) lifts automatically at 3+ races of current-season data. See the calibration reference for the shared pipeline.

← All models