The driver’s contribution
Performance attributed to the driver, with the influence of car and conditions accounted for.
ENTRANCE / POWER
Power Rankings
Loading the latest published release…
The machinery
The car’s performance, supported by comparable samples and a defensible treatment of reliability.
The combination
The driver and car evaluated together in one model, with an explicit scale and uncertainty.
Power Score · Driver + Car = Power · 100 = neutral reference · Higher is stronger · No upper cap
Measured pace, Grand Prix by Grand Prix
How the ranking works
Reading the Power Score
Driver = 50 + 10 × driver effect. Car = 50 + 10 × car effect. Power = Driver + Car = 100 + 10 × (driver effect + car effect). A score of 100 is the neutral reference, not a maximum. Higher is stronger, with no upper cap or per-release rescaling.
The underlying model estimates conditional dry race pace and separates driver and car effects through historical connections, comparable samples and explicit priors. The public score is a fixed display conversion of those effects, so it does not change the model’s ranking order. It is not a win probability or a measure of pure talent.
Qualifying, reliability and incident responsibility are not scored. Unsupported estimates remain provisional or unranked rather than receiving invented precision.
Evidence and changes
A Not rated cell opens the exact reason. One GP is one update; evidence and 90% model-conditional uncertainty are shown on the same public score scale.
Every release records its model, score contract, cutoff, coverage and uncertainty. Driver and Car separation depends on explicit priors and the historical driver/team graph.
Global ranks stay fixed when filtering. Position changes appear only between comparable sporting releases. Methodology, source-correction and producer-contract changes are not sporting gains.