> ## Documentation Index
> Fetch the complete documentation index at: https://offthepace.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Who is genuinely faster in the wet?

> Each driver's wet-weather skill advantage their race skill residual in the wet minus their dry baseline across their full 2018–2024 career.

Rain reshuffles the order, but how much of that is the driver and how much is the car's wet setup? This chart isolates the driver: it compares each driver's race skill residual on wet-flagged races against their own dry-race baseline, across their full career. Because the residual already removes car pace, tyres, fuel, and conditions, a positive wet advantage reflects genuine adaptability rather than a good wet car.

## What it shows

Each driver carries a single wet-advantage figure:

* **Wet advantage**: wet skill residual minus dry skill residual. Positive = the driver is faster (relative to the field model) in wet conditions than their dry-race baseline. It is shrunk toward zero by sample size (a k=6 prior), so low-sample drivers appear closer to neutral.
* **Wet flag (`race_wet_flag`)**: from `fct_driver_skill_features`; marks a race where weather significantly affected conditions during the timed stint window.
* **Eligibility**: at least 4 wet races and 5 dry races, each with ≥5 clean laps. Drivers below that are excluded.

<Note>
  The skill residual already accounts for car pace, tyre compound, fuel load, and ambient conditions so the wet advantage reflects driver adaptability, not the car's wet-weather setup.
</Note>

## How to use it

1. **Read the ranked wet-advantage bars** drivers at the top consistently outperform their dry baseline when it rains.
2. **Remember the shrinkage** a driver near zero may simply have few wet races, not zero wet skill.
3. **Cross-check eligibility** only drivers meeting the wet/dry minimums appear; absence is a sample-size statement.
4. **Export the data** to compare wet advantage against career length or era.

## Reading the signal correctly

This is a **within-driver** difference (wet minus that driver's own dry baseline), so it is not a wet-weather pace ranking a slower driver with a large positive advantage gains more *relative to themselves* in the rain than a quick driver who is merely consistent.

<Warning>
  Wet races are rare and heterogeneous (a damp start versus a full monsoon are both "wet"). Even after shrinkage, treat wet advantage as a lower-confidence signal than dry-season ratings, and read it alongside the [decomposition limitations](/decomposition/limitations).
</Warning>

## Data source

The chart queries `fct_driver_skill_features` via DuckDB-Wasm in the browser, aggregating across each driver's career.

| Column                      | Meaning                                                |
| --------------------------- | ------------------------------------------------------ |
| `driver_skill_proxy_mean_s` | Per-race skill residual, split by wet/dry flag         |
| `race_wet_flag`             | Whether weather materially affected the race           |
| `wet_advantage_s`           | Wet residual minus dry residual, shrunk by sample size |

The data window covers a **career aggregation over 2018–2024**.

For the full model definition, see the [`fct_driver_skill_features` reference](/reference/models/fct/fct_driver_skill_features).