Models / LightGBM (3-class)

Pit Stop Count

P(one / two / three+ stops) for the whole race, per driver.

LightGBM (3-class)analytics.strategy

How it works

Race-level strategy shape from pre-race features — circuit degradation profile, pit-lane loss, grid slot, race distance, starting compound, and the rain signal. Features are race-static apart from weather, so the distribution mostly moves with conditions.

Inputs

  • circuit deg profile
  • pit-lane delta
  • grid position
  • race distance
  • rain probability
  • starting compound

Output

drivers[].pitStopCountpOneStop / pTwoStop / pThreePlusStop / expectedStops.
← All models