Next-round ET Prediction
Uses your own logged runs for the selected vehicle + track. Blends a stable regression with nearest-neighbor matching and returns an uncertainty band.
How prediction works (transparent)
- Regression: ridge-regularized linear model of ET vs (DA, Temp, Humidity, Pressure, Grains). Stable on noisy data.
- Nearest neighbors: finds the most similar historical conditions and averages their ET deltas.
- Blend: weighted by how much data you have (few runs → more kNN; many runs → more regression).
- Uncertainty: 1-sigma band from residual spread of your matched data.
Quick Stats (Vehicle + Track)
Slip Photo Import (OCR)
Upload a timeslip photo and auto-extract RT/60/330/ET/MPH. You confirm before saving.
OCR is best with clear, straight photos.