Rain expected — wet weather specialists may outperform their qualifying position.
Predicted using an XGBoost model trained on 2018-2024 qualifying sessions. Key signals: constructor championship standing, recent 3-session qualifying form, circuit-specific history, and momentum trend. Average accuracy: 7.6/10 correct per session.
Predicted using an XGBoost model trained on 7 seasons of race data. Key signals: qualifying position, championship standing, recent 3-race form, circuit history, Google Trends momentum, and weather conditions. Average accuracy: 7.9/10 correct per race.
Constructor score combines both drivers predicted qualifying probabilities. A team with two drivers predicted in the top 5 scores highest. Weighted by championship position and recent team form.
Edge = model probability minus implied market probability. Market odds are based on qualifying position using exponential decay — how bookmakers typically open F1 lines. A positive edge means the model sees more value than the market implies. Backtested at 60% win rate at 20%+ edge threshold over 2020-2024.
All drivers ranked by race model probability. Green edge = model rates driver higher than market. Drivers where model and market agree strongly are lower edge opportunities.