fct_lap_residuals pipeline that runs across every race from 2018 to 2025, and every finding must pass the CI-enforced additive identity before it is published.
What an attributed finding is
The decomposition separates each lap’s gap to the field into fuel load, tyre compound age, constructor pace, dirty-air penalty, and driver skill residual. On any single race weekend, those terms tell you exactly what a driver contributed versus what the car, the tyres, and the strategy handed them. An attributed finding applies that decomposition to a specific race, isolates the key strategic or competitive moment, and reports the result in seconds not adjectives. The methodology behind every term is documented in the Additive Identity reference; the findings here are where that methodology meets real race outcomes.Why these matter
Lap time alone is a blunt instrument. Two drivers can post identical lap times for completely different reasons: one fighting degraded tyres while the other cruises on fresh rubber, one extracting every tenth in clean air while the other loses pace in a dirty wake. Without decomposition, a commentator watching those numbers scroll by sees parity. The model sees the opposing forces that cancel each other out and makes them visible. Attributed findings demonstrate that the framework produces meaningful, non-trivial insights on actual race data not a theoretical construction that looks clean in a notebook but dissolves when applied to the noise of a real grand prix. At São Paulo 2021 it separates the car, tyre age and traffic from a residual; it does not separate driver from tyre wear inside that residual.The primary case study: Hamilton vs Verstappen, São Paulo 2021
The São Paulo 2021 Brazilian Grand Prix is the first and flagship attributed finding. Lewis Hamilton started tenth after a grid penalty and won by 10.5 seconds. Commentary called it brilliant driving. The decomposition measures how much of the margin the car, the tyres and the traffic explain. The key finding: over the last 13 laps Hamilton gained 11.0 s on Verstappen. The car (1.94 s) and the three-lap-younger tyres (0.86 s) explain about a quarter of that, and the dirty-air tax cost Hamilton 0.33 s. The remaining 8.54 s sits in the driver skill residual, which the model cannot split between Hamilton’s driving and Verstappen’s tyres falling away faster than the model expects.São Paulo 2021: where Hamilton's 10.5 seconds came from
Hamilton wins from tenth by 10.5 seconds. The decomposition prices the car, the tyre age and the traffic in seconds and shows that most of the late-race margin is residual.